Schemas
CreateAgentContextBody
titletypesortOrderUpdateAgentContextBody
typetitlesortOrderCreateAgentDefinitionBody
slug^[a-z0-9]+(-[a-z0-9]… · requiredUnique per workspace. Immutable after creation — it IS the agent_id.
nameinstructionsThe agent's SYSTEM PROMPT — who it is and how it works. Per-run data arrives as the invocation's input and context, never from here.
enabledtoolsTool registry keys. Validated against the live catalog on write.
toolsRequiringApprovalAgentDefinitionModel
tierModel capability class — quick, standard, top or ultra. Ignored when id is set.
idExplicit model id for the active LLM provider; a Bedrock-style id is normalized on the Anthropic API. Beats tier.
temperaturemaxTokensthinkingEnabledthinkingEffortpromptCachingAgentDefinitionMcpServer
Decision Table
| Variant | Matching Criteria |
|---|---|
| type = object · kind="productmind" · requires: id | |
| type = object · kind="integration" · requires: id, instanceId | |
| type = object · kind="http" · requires: id, url |
idIdentifier for this server within the definition. Never used as a shared cache key.
kindallowedToolsTool-name globs to load from this server. Required for the asynchronous path — an absent list means 'everything this server exposes', which cannot be validated at publish time.
rejectedToolsTool-name globs to exclude, applied after allowedTools and on top of the global rejections.
ProductmindMcpServer
idIdentifier for this server within the definition. Never used as a shared cache key.
kindallowedToolsTool-name globs to load from this server. Required for the asynchronous path — an absent list means 'everything this server exposes', which cannot be validated at publish time.
rejectedToolsTool-name globs to exclude, applied after allowedTools and on top of the global rejections.
IntegrationMcpServer
idIdentifier for this server within the definition. Never used as a shared cache key.
kindinstanceIdWorkforce instance id of an MCP integration already attached to this workspace.
allowedToolsTool-name globs to load from this server. Required for the asynchronous path — an absent list means 'everything this server exposes', which cannot be validated at publish time.
rejectedToolsTool-name globs to exclude, applied after allowedTools and on top of the global rejections.
HttpMcpServer
idIdentifier for this server within the definition. Never used as a shared cache key.
kindurlThe MCP server endpoint. Our infrastructure connects outbound to it, unauthenticated.
allowedToolsTool-name globs to load from this server. Required for the asynchronous path — an absent list means 'everything this server exposes', which cannot be validated at publish time.
rejectedToolsTool-name globs to exclude, applied after allowedTools and on top of the global rejections.
transportAgentDefinitionSettings
maxTurnsCap on agent-loop turns. The bound on how long a tool-using run can iterate.
maxTotalTokensCap on tokens processed per run, including cache reads and writes; not cost-weighted.
UpdateAgentDefinitionBody
nameinstructionsThe agent's SYSTEM PROMPT — who it is and how it works. Per-run data arrives as the invocation's input and context, never from here.
enabledtoolsTool registry keys. Validated against the live catalog on write.
toolsRequiringApprovalversionExecuteAgentDefinitionBody
inputThe per-run instruction. Leads the user message. Omit it for an agent whose instructions already describe the whole job.
What caused this run. Rendered after input. Defaults to a by-hand trigger.
AgentTrigger
Decision Table
| Variant | Matching Criteria |
|---|---|
| type = object · requires: entityType, entityId, event | |
| type = object · requires: kind |
entityTypeWhich entity the event happened to.
entityIdThe entity's id — never the entity itself.
eventPast tense: the fact that already happened.
kindchangedFieldsWhich of the entity's fields this event changed, named as the API spells them.
AgentEntityTrigger
entityTypeWhich entity the event happened to.
entityIdThe entity's id — never the entity itself.
eventPast tense: the fact that already happened.
kindchangedFieldsWhich of the entity's fields this event changed, named as the API spells them.
AgentWokenTrigger
schedule for a cron tick, manual for a by-hand run. Neither carries an entity.
automationThe rule's name, as prose for the model. Absent when no automation is behind the run.
cadencePlain-language schedule, e.g. Every day at 09:00 (UTC). Present whenever the rule has one, including a by-hand test of a scheduled rule.
CreateAutomationBody
namerunAsUserId^([0-9a-fA-F]{8}-[0-…Defaults to the calling user.
AutomationTrigger
Decision Table
| Variant | Matching Criteria |
|---|---|
| type = object · requires: type, eventType | |
| type = object · requires: type, frequency, timezone | |
| type = object · requires: type |
typeeventTypeA workspace event this automation reacts to, e.g. bug_report.created.
AutomationConditions
combinatorAutomationConditionNode
Decision Table
| Variant | Matching Criteria |
|---|---|
| type = object · requires: field, operator | |
| type = object · requires: combinator, children |
fieldA field declared for this trigger's event by automationConditionFields.
operatorAutomationCondition
fieldA field declared for this trigger's event by automationConditionFields.
operatorAutomationConditionGroup
combinatorAutomationAction
Decision Table
| Variant | Matching Criteria |
|---|---|
| type = object · requires: type, agentSlug, input |
typeagentSlug^[a-z0-9]+(-[a-z0-9]… · requiredThe agent definition to run. Must resolve, be enabled and be async-eligible.
inputThe per-run instruction, as a template over {{entity.id}}, {{entity.key}}, {{entity.title}}, {{entity.status}}, {{event}}, {{actor.name}} and {{changes.<field>.to}} / .from. Unknown placeholders render empty and are recorded as warnings on the run.
AutomationLimits
maxRunsPerHourRuns per hour for this automation, clamped to [1, 200]. Defaults to 50. Counted with an indexed query, and it FAILS CLOSED.
UpdateAutomationBody
nameenabledrunAsUserId^([0-9a-fA-F]{8}-[0-…RunAutomationBody
entityId^([0-9a-fA-F]{8}-[0-…entityTypechangedFieldsinputforceCheckAutomationBody
entityId^([0-9a-fA-F]{8}-[0-…entityTypechangedFieldsCreateCommentBody
entityentityIdcontentkindOmit for a plain comment. A coding agent replying to a comment passes agent_response. agent_action and system are written by Productmind itself and refused here.
parentCommentIdcodeAccessApp only - omit. On a comment that asks PM (@pm) on a feature or component specification: true has a read-only coding agent answer instead, from the feature's code, as a coding run.
CommentAnchor
Decision Table
| Variant | Matching Criteria |
|---|---|
| type = object · requires: type, startNodeKey, startOffset +3 more | |
| type = object · requires: type, pullRequestUrl, pullRequestNumber +4 more | |
| type = object · requires: type, port, pagePath +4 more |
typestartNodeKeystartOffsetendNodeKeyendOffsetsnippetprefixsuffixtextStarttextEndEditDocumentPassageBody
summaryOne line on what you changed, shown with the edit in the thread.
The change, as exact find-and-replace edits on the highlighted blocks' current text (the markdown your turn input shows under "The highlighted blocks"), applied in order, all or none. Each find is copied exactly (same characters, spacing and line breaks) from those blocks and occurs exactly once in them: include enough surrounding text to make it unique. replace is the new text. Only those blocks can change. Prefer this over content.
contentThe highlighted blocks' WHOLE new markdown, only when the thread asks to rewrite all of them. It replaces those blocks, never anything else. Never send it with edits.
CreateDecisionBody
entityentityIddecisionTypeoutcometitlerationaleconditionalNotesdecidedByUserIddecidedByNamesupersedesIdcitedInterestIdsRequestPolymorphicAttachmentUploadBody
entityentityIdfileNamecontentTypefileSizedescriptionCreateWorkspaceInput
namedescriptionCreateFeatureBody
titleA short sentence-case noun phrase in the vocabulary a product manager and a developer both use unchanged — the domain concept, plus the operation nominalised onto it where the concept alone does not name the capability. Password reset, not Resetting a forgotten password; Customer management, not Managing customers and not a bare Customer. No articles, possessives or imperatives, and no implementation words (service, handler, worker, API) unless the product's own user-facing wording uses them. The same register at every level: a parent names a broader concept, not a different kind of word.
parentFeatureIdkindcomponent for a unit that owns code and therefore owns business rules — a screen, a service module, a background worker, or the shared code a domain's capabilities have in common. feature for the business capability above it, whose description is composed from its components. domain ONLY for a named collection that organises features: it is purely organisational, owns no code, composes nothing and is never scanned. A person makes one; a catalogue scan makes one only when its instructions say the workspace's catalogue shape is domains, features and capabilities. In every other case a scan must never mint one, and when cataloguing you should treat every container as a feature. A child may not be broader than its parent (nothing above a component, no domain under a feature). When in doubt, ask whether a deep read of this row alone would have specific files to open — if not, it is a feature.
descriptionMarkdown, at most 200,000 characters.
priorityrequestingAccountIdrequestingContactIdownerIdAssign ONE user by UUID (legacy single-assignee field, historically the feature's 'owner'). Use assignees to assign several people or a team.
REPLACES the item's whole assignee set: every user and team listed becomes an assignee, everything not listed is unassigned. Pass [] to unassign everyone. When both this and the legacy single-user field (assigneeId / ownerId) are supplied, this wins. Find a team's id with list-teams and a person's with list-assignable-users.
estimatedEfforttargetDatesortOrderpositionWhere to place the feature among its siblings (same parent) when sortOrder is not supplied. end appends it below the last sibling, start puts it above the first. Ignored when an explicit sortOrder is given.
externalSourceexternalIdexternalSourceUrlAssignmentTargetInput
entityTypeuser or team.
id^([0-9a-fA-F]{8}-[0-… · requiredThe user's or the team's UUID.
FindFeaturesByCodePathsBody
One entry per repository you changed. One call covers every repository.
BulkCreateFeaturesBody
UpdateFeatureStatusInput
statusInlineStatusDecision
outcometitlerationaleconditionalNotessupersedesIdcitedInterestIdsReorderFeatureInput
sortOrderPosition among the new siblings, 0-based. The server re-indexes the rest of that sibling list around it, and re-indexes the old siblings too when the parent changes.
Where the feature lands: a feature id nests it under that parent, null promotes it to the top level, and omitting the key leaves the current parent alone. A move that would make a feature its own ancestor is rejected.
UpdateFeatureInput
kindcomponent for a unit that owns code and therefore owns business rules — a screen, a service module, a background worker, or the shared code a domain's capabilities have in common. feature for the business capability above it, whose description is composed from its components. domain ONLY for a named collection that organises features: it is purely organisational, owns no code, composes nothing and is never scanned. A person makes one; a catalogue scan makes one only when its instructions say the workspace's catalogue shape is domains, features and capabilities. In every other case a scan must never mint one, and when cataloguing you should treat every container as a feature. A child may not be broader than its parent (nothing above a component, no domain under a feature). When in doubt, ask whether a deep read of this row alone would have specific files to open — if not, it is a feature.
Which plane a COMPONENT's code sits in. Derived from the node's FILES by a catalogue run — do not guess it: leave the field out unless you are mirroring a scan's own classification or a human's explicit choice. Null clears it, which is correct for a container that owns no code and for any feature (a feature spans layers and carries no layer, ever).
titleA short sentence-case noun phrase in the vocabulary a product manager and a developer both use unchanged — the domain concept, plus the operation nominalised onto it where the concept alone does not name the capability. Password reset, not Resetting a forgotten password; Customer management, not Managing customers and not a bare Customer. No articles, possessives or imperatives, and no implementation words (service, handler, worker, API) unless the product's own user-facing wording uses them. The same register at every level: a parent names a broader concept, not a different kind of word.
descriptionMarkdown, at most 200,000 characters. Replaces the whole description: send it only with the complete new text. To change part of the description, send descriptionEdits instead.
Change PART of the description without resending it: a list of exact find-and-replace edits the server applies to the stored description, in order, all or none. Each find must be copied exactly (same characters, spacing and line breaks) from the description as it stands, and must occur exactly once there at that point: include enough surrounding text to make it unique. replace is the new text (an empty string deletes the passage). Everything outside the passages is kept as it is. Prefer this over description for any change to a long description, and always when your read of the description was cut. Cannot be sent together with description.
prioritystatusSet the single USER assignee (legacy 'owner' field): replaces the feature's user assignees with this one, null unassigns every user. Team assignees are left untouched — use assignees to change them.
REPLACES the item's whole assignee set: every user and team listed becomes an assignee, everything not listed is unassigned. Pass [] to unassign everyone. When both this and the legacy single-user field (assigneeId / ownerId) are supplied, this wins. Find a team's id with list-teams and a person's with list-assignable-users.
sortOrderLinkFeatureApplicationBody
applicationIdcodePathsRepository-relative paths that implement this feature, most important first: entry-point files, service/domain logic, validators and guards, the constants holding literal values, the data model, UI surfaces and the tests that pin the behaviour. A directory is fine when the whole directory implements the capability; narrow a shared file with a symbol (src/routes/api.ts#createBooking). Five to forty entries is normal — src/ alone is worse than useless. Passing this key REPLACES the stored list; omitting it leaves the stored list untouched.
entryPointsExactly the entry points attributed to this feature, one per array entry, each as "POST /bookings — src/booking/booking.routes.ts:42, cron: reap-previews — src/jobs/reaper.ts:18). A path is not an entry point: the deep dive checks that every entry point listed here appears as its own contract block in the spec it writes. Passing this key REPLACES the stored list.
layerWhich layer this feature occupies IN THIS application — frontend, backend, data or external. Set by the workspace feature map, which is the only pass that sees several repositories at once and can therefore say a feature starts in a web app and ends in a database. Passing this key REPLACES the stored value; omitting it leaves it untouched, so a plain UI link can never blank it.
capabilityNotesOne to three sentences the deep-dive agent reads before it starts: whether this is a domain parent and which cross-cutting concerns it owns, where the rule-dense code is, which feature it is easily confused with, and anything surprising noticed but not pursued. A handover note, not a summary — do not repeat the description here.
FeatureApplicationManualSurfaceBody
manualCodePathsmanualEntryPointshiddenCodePathsmanualSystemComponentIdshiddenSystemComponentIdsLinkFeatureDependencyBody
targetFeatureIdThe feature this one builds on — the direction is <this feature> uses <target>: it calls the target's API, consumes a queue the target fills, or reads data the target owns. Mutual coupling is two separate calls, one in each direction. Linking an already-linked pair is safe and REPLACES the stored rationale/evidence.
rationaleOne sentence naming the MECHANISM of the dependency ("reads the bookings table BookingLedger owns", "calls POST /payments on checkout") — never a restatement that the two are related. Rendered to humans curating the map.
evidenceCode locations proving the edge, each as "
SubmitBusinessRulesBody
batchId^([0-9a-fA-F]{8}-[0-… · requiredA UUID you mint for this submit wave and reuse on a retry. Re-sending the same batch replaces it rather than doubling it, which is what makes a lost response safe to repeat.
Up to 40 rules. Submit in waves as you write them, checking each wave's evidence while the files are still open, rather than accumulating the whole set for one call.
sessionId^([0-9a-fA-F]{8}-[0-…The scan session producing these rules, when there is one.
proposalId^([0-9a-fA-F]{8}-[0-…The description proposal these rules accompany, when there is one.
finalTrue on the last wave. Promotes every wave of this batch to the component's applied rule set in one step, so a partial submit never half-replaces what is live.
FeatureBusinessRuleCitationDTO
labelrepoNameshapathPromoteBusinessRulesBody
batchId^([0-9a-fA-F]{8}-[0-… · requiredThe batch to make this component's applied rule set.
ProposeFeatureDescriptionBody
proposedDescriptionMarkdownThe FULL replacement description markdown for the feature. Code citations are link tokens — path/to/file.ext:line @
summaryOne-paragraph summary of what changed and why
sessionIdThe scan session id (from the Context section of your instructions)
REQUIRED by the deep-read instructions: either the Knowledge Base document ids you created for this feature's technical architecture (API contracts, data layers, databases, queues, execution models, state managers), or an explicit reason none was needed.
commentIdWhen you answer a comment thread on this feature's document (your turn input names it): the thread's id, exactly as given. The suggestion is then linked to the thread, and applying it resolves the thread. Only the thread you were asked to answer, while you answer it; omit it otherwise.
ApplyFeatureDescriptionProposalBody
descriptiondraftVersionApproveFeatureDescriptionProposalBody
descriptiondraftVersionProposeFeatureStructureBody
parentTitleNew title for THIS feature, retitled as a high-level domain parent (e.g. 'Finance'). Omit to keep the current title. A short sentence-case noun phrase in the vocabulary a product manager and a developer both use unchanged — the domain concept, plus the operation nominalised onto it where the concept alone does not name the capability. Password reset, not Resetting a forgotten password; Customer management, not Managing customers and not a bare Customer. No articles, possessives or imperatives, and no implementation words (service, handler, worker, API) unless the product's own user-facing wording uses them. The same register at every level: a parent names a broader concept, not a different kind of word.
parentDescriptionMarkdownNew HIGH-LEVEL description for THIS feature as the domain parent — what the domain does and the sub-capabilities it contains. Omit to keep the current description.
Ordered NEW subfeatures to create under this feature, each with its own full detailed spec.
Optionally re-parent THIS feature. Pass an existing feature id to move it under that parent, or null to promote it to a ROOT feature (use this for a domain wrongly filed under an era/technology bucket). Omit the key entirely to leave it where it is — null and omitted mean different things.
summaryOne-paragraph summary of the proposed restructure and why.
sessionIdThe scan session id (from the Context section of your instructions).
ApplyFeatureStructureProposalBody
parentTitleparentDescriptionMarkdownCreateFeatureRequestBody
titledescriptionMarkdown, at most 200,000 characters.
sourceWhat KIND of request this is, e.g. customer_feedback for something customers ask for. It does not record WHO asked: pass the people in interests.
priorityestimatedEffortsortOrderOptional. Who asked for this request, recorded as interest in the same call: each entry is exactly the body add-feature-request-interest takes. A customer is { requestingContactId } (find the contact with list-contacts, or add someone not on file yet with create-contact; their account comes with them). This is what counts the demand on the request and files it under the customer's account: a name written into the description counts nothing. Every entry is checked before anything is written, so an unknown contact rejects the whole call and creates nothing. Needs permission to update feature requests. At most 20.
CreateManualInterestBody
sourceWho this interest speaks for. Defaults to customer_feedback, which needs requestingContactId. internal_stakeholders is a teammate (set stakeholderRole). competitive_intelligence, market_signal, strategic_directive and support_pattern are other input: no contact, and you are recorded as who added it. Ignored when stakeholderRole is set.
descriptionWhat they want or said, in their words where you have them, e.g. "Needs SSO before the Q3 rollout."
sentimentHow the person feels about the request, when known.
requestingContactIdThe UUID of the customer contact who asked (find it with list-contacts; each row's id is this value). Someone who is not on file yet is added with create-contact, whose id is this value too. Required for a customer interest. The contact's account is attached automatically, so there is no account field. A contact outside this workspace's organization is rejected.
positionTheir stance: supporting, neutral, opposing, or conditional (then say the condition in conditionalNotes). Mostly for teammates and other input.
conditionalNotesThe condition behind a conditional position, e.g. "Only if it ships with SSO."
stakeholderRoleSet this to record a TEAMMATE's interest instead of a customer's: the interest becomes internal_stakeholders, whatever source says. Name the teammate with stakeholderUserId or stakeholderName; with neither, it is recorded as yours.
stakeholderUserIdThe workspace member the teammate interest is from (a user UUID). Takes precedence over stakeholderName.
stakeholderNameThe teammate's name as free text, for someone who is not a workspace member.
weightOptional relative weight of this interest. Leave it unset unless the user gives one.
UpdateFeatureRequestStatusInput
statusMergeFeatureRequestsBody
sourceIdstargetIdBulkPromoteFeatureRequestsBody
idsfeatureIdBulkPromoteFeatureRequestsToTaskBody
idsFeature request UUIDs
featureIdOptional parent Feature UUID applied to every created task.
PromoteFeatureRequestToTaskBody
titleTask title. Defaults to the feature request's title.
descriptionTask description, markdown. Defaults to the feature request's description.
featureIdOptional parent Feature UUID. Omit to create a workspace-level task with no feature.
LinkInteractionInterestBody
interactionIdThe UUID of the customer meeting or conversation to record as interest (find it with list-customer-interactions). Its contact, account, sentiment and summary are copied onto the interest.
UpdateFeatureRequestInterestBody
descriptionsentimentpositionconditionalNotesweightCreateFeatureRequestRelationshipBody
targetTypetargetIdrelationshipTypeUpdateFeatureRequestInput
titledescriptionMarkdown, at most 200,000 characters. Replaces the whole description: send it only with the complete new text. To change part of the description, send descriptionEdits instead.
Change PART of the description without resending it: a list of exact find-and-replace edits the server applies to the stored description, in order, all or none. Each find must be copied exactly (same characters, spacing and line breaks) from the description as it stands, and must occur exactly once there at that point: include enough surrounding text to make it unique. replace is the new text (an empty string deletes the passage). Everything outside the passages is kept as it is. Prefer this over description for any change to a long description, and always when your read of the description was cut. Cannot be sent together with description.
prioritystatussortOrderSet the single USER assignee (legacy-shaped field): replaces the request's user assignees with this one, null unassigns every user. Team assignees are left untouched — use assignees to change them.
REPLACES the item's whole assignee set: every user and team listed becomes an assignee, everything not listed is unassigned. Pass [] to unassign everyone. When both this and the legacy single-user field (assigneeId / ownerId) are supplied, this wins. Find a team's id with list-teams and a person's with list-assignable-users.
DescribeFeatureRequestBody
sourceHow the run was triggered (auto on image upload / manual button)
ApplyFeatureRequestDescriptionProposalBody
descriptionCreateBugReportBody
titledescriptionMarkdown, at most 200,000 characters.
priorityseverityreporterTypereporterIdreporterDetailsassigneeIdAssign ONE user by UUID (the legacy single-assignee field). Use assignees to assign several people or a team.
REPLACES the item's whole assignee set: every user and team listed becomes an assignee, everything not listed is unassigned. Pass [] to unassign everyone. When both this and the legacy single-user field (assigneeId / ownerId) are supplied, this wins. Find a team's id with list-teams and a person's with list-assignable-users.
sortOrderoriginChatId^([0-9a-fA-F]{8}-[0-…The Conversation id from your context. Pass it when you are creating this from a conversation so the images the user attached travel onto the item.
featureId^([0-9a-fA-F]{8}-[0-…Optional. UUID of the Feature this bug affects (resolve a FEAT-n with search-entities). Links the new report to it in the same call, exactly as create-bug-report-relationship would: the report lands on that feature's bug list, and if it is created with no assignee it is handed to the feature's owners. An id that is not a feature in this workspace rejects the whole call and creates nothing.
applicationId^([0-9a-fA-F]{8}-[0-…Optional. UUID of the Application where the bug was observed (resolve an APP-n with search-entities). Linked in the same call. An application link never assigns anyone. An unknown id rejects the whole call and creates nothing.
MergeBugReportsBody
sourceIdstargetIdCreateBugReportReporterBody
reporterTypereporterIddetailsConvertBugReportBody
targetTypeWhat to convert this bug report into. task for tracked engineering work, feature for a roadmap item the defect revealed as a gap, feature_request for a customer demand signal. The report is closed on success, gains a system comment naming the new entity, and the new entity links back to it.
titleOverride the new entity's title. Defaults to the bug report's title.
descriptionOverride the new entity's description (markdown). Defaults to the bug report's description; for a feature_request with a blank description the title is used, because that field is required there.
priorityOverride the new entity's priority. Defaults to the bug report's priority.
CreateBugReportRelationshipBody
targetTypeWhat targetId names. feature: the product area the bug affects. This is the usual link: the report then shows on that feature's bug list and lineage, and an unassigned report is handed to the feature's owners. application: where the bug was observed. task: related engineering work.
targetIdUUID of that feature, application or task in this workspace. Resolve a FEAT-n / COMP-n / APP-n / TASK-n handle with search-entities first.
relationshipTypeIgnored: every link made here is stored as related_to. Omit it.
UpdateBugReportInput
titledescriptionMarkdown, at most 200,000 characters. Replaces the whole description: send it only with the complete new text. To change part of the description, send descriptionEdits instead.
Change PART of the description without resending it: a list of exact find-and-replace edits the server applies to the stored description, in order, all or none. Each find must be copied exactly (same characters, spacing and line breaks) from the description as it stands, and must occur exactly once there at that point: include enough surrounding text to make it unique. replace is the new text (an empty string deletes the passage). Everything outside the passages is kept as it is. Prefer this over description for any change to a long description, and always when your read of the description was cut. Cannot be sent together with description.
priorityseveritystatusSet the single USER assignee (legacy field): replaces the report's user assignees with this one, null unassigns every user. Team assignees are left untouched — use assignees to change them.
REPLACES the item's whole assignee set: every user and team listed becomes an assignee, everything not listed is unassigned. Pass [] to unassign everyone. When both this and the legacy single-user field (assigneeId / ownerId) are supplied, this wins. Find a team's id with list-teams and a person's with list-assignable-users.
tagssortOrderDescribeBugReportBody
sourceHow the run was triggered (auto on image upload / manual button)
ApplyBugReportDescriptionProposalBody
descriptionStartTaskWorkBody
taskIdsEvery task ID to start work on. If omitted, only the path :taskId is started. All taskIds must belong to the same session as the path :taskId.
SaveTaskWorkPlanBody
overallPlanMarkdown plan covering all tasks in the session. Stored as a session log entry, a message in the session conversation; it does NOT change the session description. Publish an approved plan there with update-session-description.
Per-task plan details. Each entry is stored as a task log on the corresponding task.
taskIdsTask IDs in the session. When provided, the server validates that taskPlans covers every id in this set (and only ids in this set). Pass the same taskIds you used for start-task-work.
ReportTaskProgressBody
contentProgress note, in markdown: what changed and where, anything unexpected, what comes next. Saved as a task log entry and, when the task is in a session, posted to the session's conversation.
attachmentIdsIDs of screenshots you uploaded to this session with request-attachment-upload (entity session). They show with this note in the session's conversation and in its snippets strip. Images uploaded to a task are not shown there.
SaveTaskVerificationBody
summaryVerification summary written as a session log entry. Should describe what was verified and any uploaded screenshots.
attachmentIdsIDs of attachments you uploaded with request-attachment-upload. Images uploaded to this session (entity session) show with the summary in the session's conversation and in its snippets strip; images uploaded to a task are only listed by id in the task logs.
taskIdsTask IDs in the session. When provided, the verification summary is also written as a task log on each task in the set, in addition to the session log.
CompleteTaskWorkBody
pullRequestUrlURL of the pull request that contains the implementation.
taskIdsOptional list of task IDs to mark complete. If omitted, all tasks linked to the same session as the path :taskId are used.
CreateTaskRequestBody
titledescriptionMarkdown, at most 200,000 characters.
statustypepriorityassigneeIdAssign ONE user by UUID (the legacy single-assignee field). Use assignees to assign several people or a team.
REPLACES the item's whole assignee set: every user and team listed becomes an assignee, everything not listed is unassigned. Pass [] to unassign everyone. When both this and the legacy single-user field (assigneeId / ownerId) are supplied, this wins. Find a team's id with list-teams and a person's with list-assignable-users.
creatorIdCreator user UUID (required for service/agent callers)
featureIdDeprecated alias for parentEntityType:'feature' + parentEntityId — kept for existing agents. Supplying it alongside the parentEntityType/parentEntityId pair is a 400.
parentEntityTypeThe structural parent's entity type (feature | application | session). Pair with parentEntityId — both or neither. NOTE: parentId is the parent TASK (subtask nesting), an unrelated axis.
parentEntityIdThe structural parent's UUID. Pair with parentEntityType — both or neither.
parentIdParent TASK UUID (subtask nesting) — NOT the structural parent; see parentEntityType.
startAtdueAtsortOrderpositionWhere to place the task among its siblings (same structural parent and same parent task) when sortOrder is not supplied. end appends it below the last sibling, start puts it above the first. Ignored when an explicit sortOrder is given.
externalSourceexternalIdexternalSourceUrltagsLayer tags for the task (e.g. backend, frontend, agent, test, docs, infra).
originChatId^([0-9a-fA-F]{8}-[0-…The Conversation id from your context. Pass it when you are creating this from a conversation so the images the user attached travel onto the item.
UpdateTaskBody
titleMarkdown, at most 200,000 characters. Replaces the whole description: send it only with the complete new text. To change part of the description, send descriptionEdits instead.
Change PART of the description without resending it: a list of exact find-and-replace edits the server applies to the stored description, in order, all or none. Each find must be copied exactly (same characters, spacing and line breaks) from the description as it stands, and must occur exactly once there at that point: include enough surrounding text to make it unique. replace is the new text (an empty string deletes the passage). Everything outside the passages is kept as it is. Prefer this over description for any change to a long description, and always when your read of the description was cut. Cannot be sent together with description.
statustypeprioritySet the single USER assignee (legacy field): replaces the task's user assignees with this one, null unassigns every user. Team assignees are left untouched — use assignees to change them.
REPLACES the item's whole assignee set: every user and team listed becomes an assignee, everything not listed is unassigned. Pass [] to unassign everyone. When both this and the legacy single-user field (assigneeId / ownerId) are supplied, this wins. Find a team's id with list-teams and a person's with list-assignable-users.
Deprecated alias for parentEntityType:'feature' + parentEntityId (null clears the parent). Supplying it alongside the pair is a 400.
The structural parent's entity type. Pair with parentEntityId — both set, or both null to detach.
The structural parent's UUID. Pair with parentEntityType — both set, or both null to detach.
Parent TASK UUID (subtask nesting) — NOT the structural parent.
sortOrderDescribeTaskBody
sourceHow the run was triggered (auto on image upload / manual button)
CreateAccountInput
nameThe company or organization's name, e.g. "Acme Corp". 1 to 100 characters.
accountTypeWhat the company is to you: customer, partner, internal (your own company) or other.
The company's website, as a full URL such as "https://www.example.com".
notesFree-text notes about the account, up to 2000 characters.
sourceTypeHow the record came to exist: manual, feedback_stream, research_interview or crm_import. Set when importing records; leave it out when adding one by hand.
sourceIdThe UUID of the Productmind record this one was created from, such as a feedback stream. Leave it out when adding one by hand.
externalSourceWhich CRM this record was imported from.
externalIdThe CRM's own id for this record — people/c123…, 701.
externalSourceUrlDeep link back to the record in the CRM.
isStrategicWhether this is a strategic account: one that matters especially to you, shown with a star.
estimatedArrThe account's estimated annual recurring revenue: a non-negative number with no currency symbol, e.g. 120000.
lifecycleStateWhere the account is in its relationship with you: prospect, trial, customer, expansion, at_risk or churned.
UpdateAccountInput
nameThe company or organization's name, e.g. "Acme Corp". 1 to 100 characters.
accountTypeWhat the company is to you: customer, partner, internal (your own company) or other.
The company's website, as a full URL such as "https://www.example.com". Send "" to clear it.
notesFree-text notes about the account, up to 2000 characters. Replaces the whole text: read the current notes with get-account first and send the combined text.
isStrategicWhether this is a strategic account: one that matters especially to you, shown with a star.
The account's estimated annual recurring revenue: a non-negative number with no currency symbol, e.g. 120000.
lifecycleStateWhere the account is in its relationship with you: prospect, trial, customer, expansion, at_risk or churned.
externalSourceWhich CRM this record was imported from.
externalIdThe CRM's own id for this record — people/c123…, 701.
externalSourceUrlDeep link back to the record in the CRM.
CreateContactInput
nameThe person's full name, e.g. "Jane Doe". 1 to 100 characters.
The person's email address, e.g. "jane.doe@example.com". It identifies the person: when a contact already has this address (in any letter case), that contact is returned instead and nothing is created. Leave it out when you don't know it.
titleThe person's job title or role, e.g. "Head of Operations". Up to 100 characters.
phoneThe person's phone number, as written, e.g. "+1 555 0100". Up to 50 characters.
accountIdThe UUID of the account (company) the person belongs to: an account's id from list-accounts, or from create-account for a company you just added. Leave it out only when nobody knows the company: the contact is then filed under the "Unknown" account.
isPrimaryWhether this person is the account's primary contact: the main person there, listed first and shown with a star. Marking one contact primary does not unmark another.
sourceTypeHow the record came to exist: manual, feedback_stream, research_interview or crm_import. Set when importing records; leave it out when adding one by hand.
sourceIdThe UUID of the Productmind record this one was created from, such as a feedback stream. Leave it out when adding one by hand.
externalSourceWhich CRM this record was imported from.
externalIdThe CRM's own id for this record — people/c123…, 701.
externalSourceUrlDeep link back to the record in the CRM.
UpdateContactInput
nameThe person's full name, e.g. "Jane Doe". 1 to 100 characters.
The person's email address, e.g. "jane.doe@example.com"; "" clears it. A new address is not checked against other contacts here: search list-contacts for it first, because another contact may already have it.
titleThe person's job title or role, e.g. "Head of Operations". Up to 100 characters.
phoneThe person's phone number, as written, e.g. "+1 555 0100". Up to 50 characters.
accountIdMoves the contact to another account: that account's id, from list-accounts or create-account. Every contact belongs to one account, so it cannot be cleared.
isPrimaryWhether this person is the account's primary contact: the main person there, listed first and shown with a star. Marking one contact primary does not unmark another.
externalSourceWhich CRM this record was imported from.
externalIdThe CRM's own id for this record — people/c123…, 701.
externalSourceUrlDeep link back to the record in the CRM.
releaseUpdatesOptOuttrue stops the release update emails to this person; false sends them again.
MatchContactsInput
emailsAttendee email addresses to match against existing contacts
PromoteNoteBody
targetTypeWhat to convert this note into. feature_request for a demand signal, feature for a roadmap item, task for engineering follow-up, bug_report for an observed defect. The note is archived on success and the new entity links back to it.
featureId^([0-9a-fA-F]{8}-[0-…Task conversions only: the UUID of the feature the new task belongs to, in this workspace. Omit it for a task with no feature. Sending it with any other targetType is a 400. Ignored when the note was already converted: the call returns the item it became.
CreateCustomerInteractionInput
typetitleHuman name for the interaction. Stored in the summary blob; an AI summarization may later replace it.
customerResearchIdworkspaceIdcontactIdtagsrawContentguidelinescheduledAtendsAtWhen the meeting ends. Only meaningful paired with scheduledAt, and must be strictly after it. Leave it out when the duration is unknown — the UI renders an unset end as one hour.
meetingUrlvoiceEnabledrecordingEnabledtranscriptionEnabledPeople on the meeting invitation. The server normalizes these and stores them as read-only participants; blank or malformed addresses are dropped rather than rejected.
externalSourceWhich external system this interaction was imported from.
externalIdThe external system's own id for this record — half of the dedup key.
externalSourceUrlDeep link back to the record in the external system.
UpdateCustomerInteractionInput
tagsrawContentguidelinemeetingUrlrecordingEnabledtranscriptionEnabledThe conversation's whole participant roster, replaced. The server normalizes the addresses, keeps the role it already recorded for anyone already on the roster, and drops blank or malformed entries. Send an empty array to clear it.
UpdateCustomerInteractionStatusInput
statusWhy the interaction is being cancelled, shown beside its cancelled status. Kept only with status cancelled; any other status clears it.
A machine-readable code for the cancel reason, such as insufficient_credit or spending_paused for an interview the AI credit check refused. Kept only with status cancelled.
DescribeCustomerInteractionBody
sourceHow the run was triggered. Only manual is wired today.
ApplyCustomerInteractionDescriptionProposalBody
descriptionCreateCustomerInteractionMessageInput
contentisFromAgentSet by Productmind's interviewer only; refused from any other caller.
Set by Productmind's interviewer only; refused from any other caller.
userIdSet by Productmind's interviewer only; refused from any other caller.
organizationIdSet by Productmind's interviewer only; refused from any other caller, and never read.
CreateInternalMeetingInput
titleWhat the meeting is about — a real column here, unlike CustomerInteraction where the name hides inside the summary blob.
agendaWhat the meeting intends to cover. Fed to the summarizer as steering context, so it shapes the summary as well as the invitation.
notesMeeting notes, pasted. The Recall notetaker writes a transcript into this same field.
tagsscheduledAtendsAtWhen the meeting ends. Only meaningful paired with scheduledAt, and must be strictly after it. Leave it out when the duration is unknown — the UI renders an unset end as one hour.
meetingUrlGoogle Meet / Zoom / Teams link the notetaker joins.
recordingEnabledSend the Recall notetaker to this meeting.
transcriptionEnabledPeople on the meeting invitation. The server normalizes these into participants and resolves platform users among them; blank or malformed addresses are dropped rather than rejected.
externalSourceWhich external system this meeting was imported from.
externalIdThe external system's own id for this event — half of the dedup key.
externalSourceUrlDeep link back to the event in the external system.
UpdateInternalMeetingInput
titleagendanotestagsmeetingUrlrecordingEnabledtranscriptionEnabledThe meeting's whole participant roster, replaced. The server normalizes the addresses, resolves platform users among them, keeps the role it already recorded for anyone already on the roster, and drops blank or malformed entries. Send an empty array to clear it.
InternalMeetingActionPoint
descriptionassigneeUserIdassigneeEmaildoneUpdateInternalMeetingStatusInput
statusCreateInternalMeetingRelationshipBody
targetTypefeature, feature_request or task.
targetIdUUID of the entity to link.
SupportImportPreviewBody
SupportTicketIngest
platformSupport platform the ticket came from
externalIdPlatform ticket id — half of the dedup key
externalSourceUrlDeep link back to the ticket
subjectTicket subject line
requesterEmailRequester email — what contact matching keys on
requesterNameRequester display name
statusPlatform status name (resolved, closed, open, …)
tagsTicket tags, used by the automation filters
groupIdPlatform group / team id
createdAtISO 8601 ticket creation timestamp
resolvedAtISO 8601 resolution timestamp
descriptionThe requester's opening message
Public reply thread, oldest first. Internal notes are excluded by the connector.
Files from the description and the public replies. Internal notes' files are excluded by the connector. Each carries a short-lived downloadUrl the API fetches during ingest; unparseable entries are dropped.
interactionTypeOverride the interaction type; defaults to support
contactId^([0-9a-fA-F]{8}-[0-…Link to this contact instead of the one matched from requesterEmail
SupportImportBody
instanceId^([0-9a-fA-F]{8}-[0-…summarizeSummarize each imported conversation with AI (default: true)
UpdateSupportIngestionConfigBody
statusestagstagMatchModegroupIdsdefaultInteractionTypeCreateSupportRequestScreenshots
conversationAttachmentIdsSupportScreenshotClaim
uploadId^([0-9a-fA-F]{8}-[0-… · requiredfileName^[^/\\\p{Cc}\p{Cf}\p… · requiredCreateCustomerResearchBody
titleobjectivesquestionsdescriptionwidgetWelcomeMessageUpdateCustomerResearchInput
titledescriptionobjectivesquestionsguidelineinterviewWelcomeMessagewidgetWelcomeMessageCreateKnowledgeBaseDocumentBody
titledescriptioncontentOptional initial markdown body. When set, the server derives the editable Lexical/Y.Doc state, snapshots version 1, and marks the document as generated.
parentDocumentIdOptional parent KB document id — creates this document as a child node in the document tree. System documents cannot be parents or children.
applicationIdOptional application id to auto-link this document to (kb_document documents application).
availableToAgentsShare the document with agents (chunk + embed) once content is saved. Defaults to false; only meaningful together with content.
tagsOptional content-domain tags — product (specs, overviews, personas, UX), engineering (technical documentation), research (interviews, competitor/market research). Agents filter KB search on this axis. When omitted, a document created with content defaults to ["engineering"]; providing the field (even an empty array) is honored exactly.
LinkKbDocumentBody
entityTypeTarget entity type
entityIdTarget entity UUID (application or feature id)
ProposeKbDocumentContentBody
proposedContentMarkdownThe FULL replacement markdown for the document. Code citations are link tokens — path/to/file.ext:line @
summaryOne-paragraph summary of what changed and why
sessionIdThe session id (from the Context section of your instructions)
commentIdWhen you answer a comment thread on this document (your turn input names it): the thread's id, exactly as given. The suggestion is then linked to the thread, and applying it resolves the thread. Only the thread you were asked to answer, while you answer it; omit it otherwise.
ApproveKbDocumentProposalBody
contentdraftVersionApplyKbDocumentProposalBody
contentdraftVersionProposeKbDocumentCreationBody
titleThe title the document will be created with. The root of an application's technical documentation is "
proposedContentMarkdownThe FULL markdown of the new document. Code citations are link tokens — path/to/file.ext:line @
applicationIdThe Application this document documents (its id from the Context). The created document is linked to it.
parentDocumentIdAn EXISTING document to nest the new one under. Omit for a root document. Use parentProposalId instead when the parent is another creation you proposed in this same session.
parentProposalIdThe id returned by an earlier propose-kb-document-creation call in this session whose document the new one nests under — how a child is attached beneath a root that does not exist yet. Resolved when both are applied; a dismissed parent puts the child at root.
summaryOne paragraph: why this document is needed and which changed files motivated it
sessionIdThe session id (from the Context section of your instructions)
KnowledgeBaseUpdateDocumentInput
titleRename the document. Trimmed, then 1 to 255 characters. Refused with 409: on a system document (Overview, Personas, User experience, Competitors keep their names; kb_document_title_fixed), when another document in the workspace already has exactly this title (kb_document_title_taken), and on a document synced from a repository (change its title in the repository; kb_document_repository_synced).
tagsdescriptionKnowledgeBaseSearchInput
querymodelimittagsexcludeTagsKnowledgeBaseRagInput
questionpromptTemplatemodelimittagsIngestConnectorItemsBody
connectorInstanceIdproviderIngestTextDocumentBody
namecontentchunkingStrategyuserIdIgnored. The document is attributed to the person the request acts as, never to a person this field names.
KnowledgeBaseConnectorRef
providerexternalIdexternalUrlCreateChatBody
titleinitialMessageagentIdclientChatId^([0-9a-fA-F]{8}-[0-…An id the client minted for this conversation before creating it. The chat is stored under chatIdFor(workspaceId, userId, clientChatId), so the client knows its id at once, and a repeated create returns the conversation already made.
CreateReleaseBody
nameOptional human label, e.g. "November ship" or "Checkout revamp". Normally left empty — the release names itself after its generated REL-n display id, and the release-notes draft replaces that with a title derived from the actual commits. Supply one only when you already know what this release is called.
descriptionRelease notes (markdown). Normally left empty — the prepare job drafts these from the actual commits and the user edits them afterwards.
statusapplicationIdsIDs of the Applications to release. Supplying them here links the applications AND starts the background prepare (diff production vs development, list the commits, open a pull request per application, draft the notes). Each application must have a connected git repository.
UpdateReleaseBody
nameRelease notes (markdown). Pushed into every pull-request body at ship time.
statusAddReleaseShippedItemsBody
The work to put on this release. An item already on the release is silently skipped, so this is safe to retry.
CreateReleaseDeploymentBody
environment^[a-z](?:-?[a-z0-9])… · requiredThe environment's key: development, production, or one of the workspace's own environments.
commitShapullRequestUrltagNamedeployedByIddeployedAtstatusbuildUrlnotesUpdateReleaseDeploymentBody
statusGenerateReleaseArtifactBody
typeAddReleaseApplicationsBody
applicationIdsIDs of the Applications to include in this release. Each must have a connected git repository — an application with no repository is rejected.
UpdateReleaseApplicationDescriptionBody
bodyThe new description (markdown). An empty string clears it.
baseBodyThe description the edit started from, as it was read. When the pull request's description has changed since, nothing is written and the answer is a 409 carrying the current description.
ShipReleaseBody
forceMerge every application whose pull request the provider will accept, even when checks are not green. Skips the CI gate but NOT the provider's own branch protection — a merge the provider refuses still fails and is reported per application.
releaseApplicationId^([0-9a-fA-F]{8}-[0-…Release this application only: the release-application id, as the release's applications list returns it. Only its pull request is merged and tagged, so only its repository's CI/CD runs, and the other applications are left as they are. The release stays in_progress until every application has merged. Omitted, every application not merged yet is released.
CreateMarketingReleaseBody
titleSources to attach at once; none is needed.
MarketingReleaseSourceRef
typerelease, feature, feature_request or customer_interaction (customer feedback).
id^([0-9a-fA-F]{8}-[0-… · requiredThe source's id, in this workspace.
UpdateMarketingReleaseBody
titlestatusAttachMarketingReleaseSourcesBody
RequestMarketingMediaUploadsBody
The files to make a place for, at most 100 a call. Each gets its own upload link in the answer.
The campaign these files were rendered from. Give it for a rendered campaign: its files then replace the ones delivered before under the same names. Leave it out for a single file of your own.
sessionId^([0-9a-fA-F]{8}-[0-…The coding session that rendered the files, when one did: its own session id.
MarketingMediaUploadFile
fileName^(?![./])(?!.*(?:^|\… · requiredThe file's name: its path inside the campaign's output folder, exactly as the manifest lists it (homepage-wide.mp4, banners/homepage-launch-open-graph.png), or a plain name for a single upload. With campaign, a file of the same name delivered before is replaced.
contentTypeThe file's media type. A clip is video/mp4 (or video/webm); a loop image/webp; a poster or banner image/png, image/webp or image/jpeg; captions application/x-subrip (.srt) or text/vtt.
bytesThe file's exact size in bytes. The upload link takes that many bytes and no others.
sha256^[0-9a-f]{64}$ · requiredThe file's sha256, as 64 lower-case hex digits (sha256sum <file>). It is recorded with the file, which is known by it afterwards.
kindWhat the file is: clip (a video), loop (a short animated image), poster (a clip's still frame), banner (one still at a channel's exact pixels) or captions (a narrated clip's captions file).
Which cut of the campaign the file is.
widthThe picture's width in pixels.
heightThe picture's height in pixels.
durationMsA clip's or a loop's length in milliseconds.
labelA short name a person will recognize it by. Optional.
MarketingMediaVariant
formatA clip's or a poster's format: wide, portrait or square.
themelight or dark.
targetA banner's target, as the marketing studio names it: linkedin-card, open-graph.
narratedTrue for a clip with its narration under it.
captionedTrue for a clip with its captions drawn into the picture.
scale2 for a poster drawn at twice its size.
ConfirmMarketingMediaBody
mediaIdsThe ids of the files that were uploaded, from request-marketing-media-uploads. Each is looked for in storage.
CreateMarketingNotesBody
generateWrite a first draft from the sources in the background.
acknowledgeInternalOnlyWith generate: the person chose Generate anyway on sources that hold no customer-facing feature.
SaveMarketingNotesBody
headlinesummarybodyGenerateMarketingNotesBody
confirmOverwriteRequired when the document has text: the person confirmed replacing it.
acknowledgeInternalOnlyThe person chose Generate anyway on sources that hold no customer-facing feature.
PublishMarketingNotesBody
expectedUpdatedAtThe draft's updatedAt the person reviewed; a newer draft refuses the publish.
quotesReviewedThe person confirmed every customer quote is anonymized. Required when the notes quote customers.
UnpublishMarketingNotesBody
versionThe newest published version the person confirmed unpublishing.
RepublishMarketingNotesBody
versionThe newest published version the person confirmed making live again. No new version is created.
CreateBlogPostBody
acknowledgeMissingSpecificationsThe person chose Generate anyway with a feature in scope that has no specification.
BlogPostScopeInput
featureIdsCustomer-facing features of the Marketing release's sources that the post is about.
The published notes document the passages were picked from. Required with passages.
The published customer notes version the passages were picked from. Required with passages.
Whole sections (0 is the summary) or highlighted excerpts of the published customer notes.
BlogPostGuidance
tonelengthangleA free-text angle for the post, e.g. who it is for or what to lead with.
GenerateBlogPostBody
A new scope; the post's own when omitted.
confirmOverwriteRequired when the post has text: the person confirmed replacing it.
acknowledgeMissingSpecificationsThe person chose Generate anyway with a feature in scope that has no specification.
UpdateBlogPostScopeBody
ApproveBlogPostBody
expectedUpdatedAtThe post's updatedAt the person reviewed; a newer post refuses the approval.
quotesReviewedThe person confirmed every customer quote is anonymized. Required when the post quotes customers.
PublishBlogPostBody
summaryWhat the blog's index and link previews say of the post.
expectedUpdatedAtThe post's updatedAt the person reviewed; a newer post refuses the publish.
slugThe post's address on the blog. Required on its first publish, then fixed: a later publish may leave it out or send the same one.
A ready image, loop or clip of the post's Marketing release's media shown at the top of the post; none when left out.
Who the post is by, as the blog shows it; none when left out or empty.
GenerateCustomerEmailsBody
notesId^([0-9a-fA-F]{8}-[0-…A live notes document of the Marketing release; the most recently published one when omitted.
SendCustomerEmailsBody
namesReviewedThe person confirmed no draft identifies another customer. Required when a draft may name one.
ProposeMarketingDocumentContentBody
summaryOne line on what the suggestion changes, shown above the review.
commentIdThe id of the comment thread you are answering, exactly as your turn input names it. Required: a suggestion is only taken from the thread you were asked to answer, while you answer it. Applying it resolves the thread.
The change, as exact find-and-replace edits on the document's current text, applied in order, all or none. Each find is copied exactly (same characters, spacing and line breaks) from the text get-marketing-document returns, and occurs exactly once there: include enough surrounding text to make it unique. replace is the new text. Change only the passage the thread is about; every other passage stays as it is. Prefer this over content.
contentThe WHOLE new document, only when the thread asks to rewrite all of it. Never send it with edits.
InsertMarketingMediaImageBody
mediaId^([0-9a-fA-F]{8}-[0-… · requiredThe media file to put into the document: a ready loop, poster or banner of the same Marketing release.
ApplyMarketingDocumentProposalBody
contentThe document's new text, as composed in the review.
CreateSandboxBody
namedescriptionsizeSandbox size: small (2 vCPU, 8 GB), medium (4 vCPU, 16 GB) or large (8 vCPU, 32 GB)
isDefaultSandboxSize
Sandbox size: small (2 vCPU, 8 GB), medium (4 vCPU, 16 GB) or large (8 vCPU, 32 GB)
SidecarEnv
key^[A-Za-z_][A-Za-z0-9… · requiredcommandWorkspaceEnvVar
key^[A-Za-z_][A-Za-z0-9… · requiredvaluesecretUpdateSandboxBody
namesizeSandbox size: small (2 vCPU, 8 GB), medium (4 vCPU, 16 GB) or large (8 vCPU, 32 GB)
isDefaultSetSandboxApplicationsBody
applicationIdsEvery application the sandbox runs, by id (from list-applications), in startup order: the first starts first. Replaces the whole list, so send the ones to keep too; [] detaches them all. New sessions on the sandbox take these applications' repositories, and the first one's base branch is what their pull requests merge into. At most 50.
UpdateSandboxSizeBody
sizesmall (2 vCPU, 8 GB), medium (4 vCPU, 16 GB) or large (8 vCPU, 32 GB). Applies from the sandbox's next hosted run, which starts on a new machine and continues from the work the coding agent pushed: a run already going keeps its size. Sandbox time is billed at the size's own hourly rate.
WorkspaceEnvVarPatchBody
Variables to add, or to replace by key, each { key, value?, secret? }. A key already stored is replaced in place; a new one is added at the end. Every key once.
removeKeys to delete. A key that is not set is ignored. A key cannot be in both set and remove.
WorkspaceEnvVarPatchEntry
key^[A-Za-z_][A-Za-z0-9… · requiredThe variable's name: letters, digits and _, not starting with a digit. Case-sensitive.
valueThe value. On a secret, leave it out (or send "") to keep the value already stored. On a plain variable, leaving it out keeps its current value, and a new variable gets "".
secrettrue stores the value encrypted: it is never shown again, by any tool or in the app. Leave it out to keep the variable's current kind; a new variable is plain unless this is true.
CreateApplicationBody
namedescriptionrequireDependencyInstallreadOnlyllmAssistedPreviewSetupApplicationRepo
workforceConnectorIdexternalRepoIdrepoNamerepoUrlproviderbaseBranchPreviewPort
portnameprotocolisMainApplicationService
nameThe service's name, e.g. web or api
typeWhat kind of unit the service is
portThe port it serves on in a live preview; leave it out for a service with nothing to preview
protocolHow the port is reached: http (the default) or https
isMainWhether this port is the one a live preview opens first
The service's OpenAPI or Swagger file, a path in the repository such as services/api/openapi.json. Leave it out to keep the stored one; null clears it, so the file is found automatically again
MonorepoService
nametypeCreateApplicationFromTemplateBody
templateId^([0-9a-fA-F]{8}-[0-… · requirednameApplicationRepairedCommandsBody
previewId^([0-9a-fA-F]{8}-[0-… · requiredThe live preview whose repaired commands to save
ApplicationEnvironmentBranchBody
The branch whose merges deploy to this custom environment, or null to unmap it.
UpdateApplicationBody
nameThe whole risk profile, replacing the stored one; null resets it to the defaults
requireDependencyInstallreadOnlyllmAssistedPreviewSetupApplicationRiskRule
idUnique within the profile
nameShown as the reason a file is flagged
tierhigh and medium flag a file; a file matching any low rule is never flagged
pathPatternsGlobs over the repository-relative path (*, **, ?, {a,b}), case-insensitive. A pattern with no / matches the file name in any folder
nameWordsWhole words of a folder or file name, case-insensitive. Names split on punctuation and camelCase, and two adjacent pieces also match joined
lineTextsText an added or removed line contains as whole words, case-insensitive, with any run of spaces between words
Matches when additions plus deletions reach this; null for no threshold
UpdateApplicationCommandsBody
The shell command that installs dependencies, run in the working directory before the dev command, e.g. bun install. null clears it; leave it out to keep it.
The shell command that starts the application for a live preview, e.g. bun run dev. null clears it; leave it out to keep it.
The directory both commands run in, relative to the repository root, e.g. services/web. null runs them at the root; leave it out to keep it.
requireDependencyInstallRequire dependencies to install: true makes a plan, a build or a scan of this application stop before the AI starts when its install command fails, times out or cannot run; false lets the run continue without it, as by default. true needs an install command, the stored one or installCommand in this call; clearing the install command turns it off. Leave it out to keep it.
UpdateApplicationServicesBody
Every runnable unit of the application, replacing the stored list: send the services to keep too, [] removes them all. Each is { name, type, port?, protocol?, isMain? }. A service with a port is a preview port: ports are unique, and at most 20 services have one. isMain marks the port a live preview opens first: one service at most, and only one with a port; when none is marked, the first service with a port is the main one. At most 50 services.
UpdateApplicationRiskRulesBody
The rules pull request reviews rate each changed file with, replacing the whole profile, or null to put the application back on Productmind's default rules. An array is the complete profile: a rule left out is no longer applied, the defaults included. get-application answers riskRules: null while the application is on the defaults. Each rule is { id, name, tier, pathPatterns, nameWords, lineTexts, minChangedLines }, every key present (empty lists and null where unused), with at least one condition.
UpdateApplicationBranchesBody
baseBranchThe branch coding sessions branch from and open their pull requests against.
The branch whose merges deploy this application to Development. null falls back to the workspace's Git workflow default.
The branch whose merges deploy this application to Production, and what a release merges into. null falls back to the workspace's Git workflow default.
Custom environments' branches, each { environmentId, branch }. Every environment once.
readOnlyRead-only: true makes coding sessions read this application's repository for context and never push to it or open pull requests on it, in every session, the open ones included; false lets sessions push their working branch again, as by default. Leave it out to keep it.
ApplicationEnvironmentBranchEntry
environmentId^([0-9a-fA-F]{8}-[0-… · requiredA custom environment's id, from list-application-environments. Not Development or Production: they take developmentBranch and productionBranch.
The branch whose merges deploy this application to the environment, or null to unmap it.
CreateEnvironmentBody
key^[a-z](?:-?[a-z0-9])… · requiredThe environment's key, which never changes.
nameThe environment's display name.
tierThe environment's tier: production or non_production.
UpdateEnvironmentBody
nameThe environment's display name.
tierA custom environment's new tier. Development and Production keep theirs.
TagFeatureMapDeltasBody
deltaIdsThe changes to tag. A tag passes to the changes under each of them; merges and removals take none.
addTags to put on the changes, spelled the standard way (lowercase-with-hyphens) when saved.
removeTags to take off the changes, and so off the changes under them.
AssignFeatureMapDeltasBody
deltaIdsThe changes to assign. The changes under each of them follow it, unless a person assigned them.
The people and teams to assign, replacing who the changes are assigned to. [] assigns nobody; null puts the changes back to automatic (their features' owners).
StartFeatureMapBody
applicationIdskbDocumentIdsKnowledge-base documents (at most 10) whose content reaches the prompts that group the code into domains and features, in the order they should be read. Omit it to use the workspace's default documents (Settings > Feature discovery); send [] to use none for this run. Each listed document must be in the workspace and neither a system document nor being deleted, and the caller must be able to read the knowledge base.
CreateSessionWithTasksBody
taskIdsTasks to attach to the new session. A task held by a session that is still open (anything but a draft in planning, or a finished session: completed, rolled back, failed or cancelled) is refused with a 409 naming both. The first id also names the working branch (task/<slug-of-its-title>-<timestamp>).
nameHuman-readable session name (e.g. 'Ticket dashboard v1'). Omit to have it derived from the task titles. The branch name is derived server-side either way; do not pass a branch name.
descriptionThe session's description, in markdown: the plan document a person opening the session reads, and the brief a coding agent drafts its plan from. Nothing fills it in for you: it is not copied from the conversation or derived from the tasks, so a session created without it has an empty description. When a plan or approach was agreed, pass it here in full (goal, approach, decisions, constraints, how to tell it is done), opening with a one-line summary. With originChatId this is the only way to hand it over, because the plan is drafted during this call.
originChatId^([0-9a-fA-F]{8}-[0-…The assistant conversation this session is being created FROM — pass the Conversation id from your context whenever you propose a session inside a conversation. Links the session to it (the conversation becomes the coding agent's context and is latched to the session) and drafts the plan immediately.
applicationIdsApplications whose repositories this session should work in, added on top of the ones the sandbox and the tasks already reach. Pass the ids the user pinned in the conversation when check-session-scope said the session would otherwise reach no repository. An application with no connected repository contributes nothing. Each one named here also runs in the session's sandbox, even when another sandbox lists it.
readOnlyApplicationIdsApplications whose repositories this session should READ for context only: each is attached like applicationIds, but read-only in this session, so the coding agent reads its code and nothing is ever committed, pushed or opened as a pull request on it. A repository an application already makes read-only (its Read-only setting) stays read-only in every session, whether or not it is named here. An application in both lists is read-only.
sandboxId^([0-9a-fA-F]{8}-[0-…The sandbox (the container the session runs in) to bind. Omit to use the workspace default - only pass one when check-session-scope reported no default and the user chose.
With originChatId only: where the plan drafted in this call runs, as the person chose it on the hand-off card. Omit it to let the server pick: the person's online Productmind agent, else the cloud.
SessionTransportChoice
modecloud, or codeflow_local for a Productmind agent on the person's machine
codeflowInstanceIdThe Productmind agent it runs on, required with mode set to codeflow_local
CreateBugFixSessionBody
bugReportIdsBug reports to attach to the new bug-fix session. Their related Applications' repos are auto-added on top of the default sandbox's. The first id also names the working branch (bugfix/<slug-of-its-title>-<timestamp>).
Sandbox to bind the session to. Omit to use the workspace default - only pass one when check-session-scope reported no default and the user chose.
App only - omit. Explicit repos to attach at create. When provided (even empty) it is honored exactly, overriding BOTH the sandbox-derived repos and the bug-related-Application layering; omit to derive them server-side.
descriptionThe session's description, in markdown: the plan document a person opening the session reads, and the brief a coding agent drafts its plan from. Nothing fills it in for you: it is not copied from the conversation or derived from the tasks, so a session created without it has an empty description. When a plan or approach was agreed, pass it here in full (goal, approach, decisions, constraints, how to tell it is done), opening with a one-line summary. With originChatId this is the only way to hand it over, because the plan is drafted during this call.
originChatId^([0-9a-fA-F]{8}-[0-…The assistant conversation this session is being created FROM - pass the Conversation id from your context whenever you propose a bug-fix session inside a conversation. Links the session to it (the conversation becomes the coding agent's context and is latched to the session) and drafts the fix plan immediately.
applicationIdsApplications whose repositories this session should work in, added on top of the ones the sandbox and the bug reports already reach. Pass the ids the user pinned in the conversation when check-session-scope said the session would otherwise reach no repository. Each one named here also runs in the session's sandbox, even when another sandbox lists it.
readOnlyApplicationIdsApplications whose repositories this session should READ for context only: each is attached like applicationIds, but read-only in this session, so the coding agent reads its code and nothing is ever committed, pushed or opened as a pull request on it. A repository an application already makes read-only (its Read-only setting) stays read-only in every session, whether or not it is named here. An application in both lists is read-only.
With originChatId only: where the plan drafted in this call runs, as the person chose it on the hand-off card. Omit it to let the server pick: the person's online Productmind agent, else the cloud.
AutonomousSessionRepoInput
workforceConnectorIdWorkforce connector-instance id that owns the repo
externalRepoIdVendor-side repo id (GitHub/GitLab/Bitbucket)
repoNameDisplay name (typically owner/name)
repoUrlCanonical HTTPS URL to the repository
providerProvider name (github, gitlab, bitbucket, …)
baseBranchBranch to clone and base the working branch on
workingBranchWorking branch name; auto-derived from the session id when omitted
readOnlytrue attaches the repository for context only: the coding agent reads it, and nothing is ever committed, pushed or opened as a pull request on it. Leave it out to inherit: a repository is read-only anyway when an application on it is.
StartCodeChatBody
messageThe first message the coding agent answers, written to stand on its own: what the person wants to understand or decide about the code, in their terms (for example 'How does refund approval work in the code, and could it be split out of checkout?'). The agent also reads the conversation passed as originChatId.
applicationIdsThe applications whose code the chat reads: their repositories are checked out read-only. Pass your best choice from the conversation (check-session-scope lists the workspace's applications; one whose repoName is null has no repository and cannot help): the person confirms or changes them on the hand-off card. With sandboxId, the sandbox's own applications are read too. At least one application read must have a repository.
sandboxId^([0-9a-fA-F]{8}-[0-…The sandbox the chat runs in, and the plan it may become: its own applications' code is read too. Omit it unless the person chose one; a plan drafted from the chat then runs in the workspace's default sandbox.
Where the coding agent runs, as the person chose it: omit it unless they did. The server then picks: their online Productmind agent, else the cloud.
featureId^([0-9a-fA-F]{8}-[0-…The feature or component the discussion is about, when it is about one: the chat is linked to it, its linked applications' code is checked out too, and the agent may propose corrections to its document. Omit it for a question about the code in general.
originChatId^([0-9a-fA-F]{8}-[0-…The conversation this chat continues: the Conversation id in your context. The coding agent is handed that conversation as context, and the conversation carries on with the coding agent. It stays private to the person. Only the conversation's owner may pass it.
attachmentIdsApp only - omit. Images uploaded by the composer against draftChatId, moved onto the new chat.
draftChatId^([0-9a-fA-F]{8}-[0-…App only - omit. The composer's draft conversation the attachmentIds were uploaded against.
clientSessionId^([0-9a-fA-F]{8}-[0-…App only - omit. An id the app minted for the new chat: the session is stored under an id derived from it, and a repeat with the same value answers the chat already started.
clientMessageId^([0-9a-fA-F]{8}-[0-…App only - omit. An id the app minted for the first message: its transcript row is stored under an id derived from it.
CreateSessionWithFeatureRequestsBody
featureRequestIdsFeature requests to launch the session from, in order. Each New or In Review request is promoted to a task; a request already promoted to a task reuses that task. The first request's task names the working branch.
The decision recorded for promoting the requests, when the workspace requires one for that transition. One decision is recorded for each request this call promotes; a request that reuses its task records none.
CreateSessionBody
namecategoryscanObjectivedescriptionThe session's description, in markdown: the plan document a person opening the session reads, and the brief a coding agent drafts its plan from. Nothing fills it in for you: it is not copied from the conversation or derived from the tasks, so a session created without it has an empty description. When a plan or approach was agreed, pass it here in full (goal, approach, decisions, constraints, how to tell it is done), opening with a one-line summary. With originChatId this is the only way to hand it over, because the plan is drafted during this call.
assigneeIdAssign ONE user by UUID (the legacy single-assignee field; defaults to the creator). Use assignees to assign several people or a team.
REPLACES the item's whole assignee set: every user and team listed becomes an assignee, everything not listed is unassigned. Pass [] to unassign everyone. When both this and the legacy single-user field (assigneeId / ownerId) are supplied, this wins. Find a team's id with list-teams and a person's with list-assignable-users.
branchNametagNameplannedDaterelatedSessionIdreleaseIdrepositoryUrloriginChatId^([0-9a-fA-F]{8}-[0-…The assistant conversation this session is being created FROM (the Conversation id in your context). Links the session to that conversation: its transcript becomes the coding agent's context, the conversation becomes team-visible and is latched to the session, and the session starts at once - a plan session drafts its plan, a live one starts its preview and sends firstMessage when it is up. Only the conversation's owner may pass it. Because the session starts at once, the create is refused (409, coding.local_agent_required) before anything is written when cloud sessions are turned off for the workspace and the person has no Productmind agent online: check-session-scope reports that first as the no_local_agent blocker.
Explicit repos to attach at create (from the quick-create repo picker). When provided (even empty) it is honored exactly and overrides the sandbox-derived repos; omit to derive repos from the sandbox.
applicationIdsApplications whose repositories this session should work in, added on top of the ones the sandbox already reaches. Pass the ids the user pinned in the conversation when check-session-scope said the session would otherwise reach no repository. An application with no connected repository contributes nothing. Each one named here also runs in the session's sandbox, even when another sandbox lists it.
readOnlyApplicationIdsApplications whose repositories this session should READ for context only: each is attached like applicationIds, but read-only in this session, so the coding agent reads its code and nothing is ever committed, pushed or opened as a pull request on it. A repository an application already makes read-only (its Read-only setting) stays read-only in every session, whether or not it is named here. An application in both lists is read-only.
featureIdsFeatures this session builds directly, when no task or bug report scopes it: the feature the user is viewing, or one they named. Links the session to each: it appears on the feature's Sessions tab and in its lineage, and the coding agent is given the feature's specification. Each must be a feature in this workspace. Omit it for a session scoped by tasks or bug reports; it reaches their features through them.
firstMessageCategory live only. The first instruction for the coding agent: what to build or change first, in the user's terms. The session's preview starts at once, and this message is sent to the agent as its first live edit the moment the preview is ready, together with this conversation as context. Required for a Live handoff; refused on any other category. Like originChatId, it starts work in this request, so the create is refused (409, coding.local_agent_required) when cloud sessions are turned off for the workspace and the person has no Productmind agent online.
startPreviewApp only, omit. Category live only. Starts the session's preview in this request with no firstMessage; held to the same checks as firstMessage.
firstMessageAttachmentIdsApp only - omit. Images for firstMessage, uploaded by the composer against its draft conversation (draftChatId) and moved onto the session.
draftChatId^([0-9a-fA-F]{8}-[0-…App only - omit. The composer's draft conversation the firstMessageAttachmentIds were uploaded against. They are moved onto the session when it holds no message, and copied otherwise.
App only - omit. Where a Live session's preview runs, as the composer resolved it. Omitted, the server picks: an online local agent, else the cloud. cloud is refused (409) while cloud sessions are turned off for the workspace.
clientSessionId^([0-9a-fA-F]{8}-[0-…App only - omit. An id the app minted for the new session: the session is stored under an id derived from it, and a repeat with the same value answers the session already made.
clientMessageId^([0-9a-fA-F]{8}-[0-…App only - omit. Category live with firstMessage only. An id the app minted for the first message: it is delivered under an id derived from it.
deferPreviewStartApp only - omit. Category live only. Hold the first message but leave the preview to a preview/start call the app makes next.
UpdateSessionBody
namecategorySet the single USER assignee (legacy field): replaces the session's user assignees with this one, null unassigns every user. Team assignees are left untouched — use assignees to change them.
REPLACES the item's whole assignee set: every user and team listed becomes an assignee, everything not listed is unassigned. Pass [] to unassign everyone. When both this and the legacy single-user field (assigneeId / ownerId) are supplied, this wins. Find a team's id with list-teams and a person's with list-assignable-users.
Rename the session's working branch. Every attached repository's working branch moves with it, so the next turn pushes the new name (use it when the git host's branch rules refused the old one). Refused (409) once a repository has a pull request or a pushed branch, while a run is in flight, and when clearing it while repositories are attached.
pullRequestUrlsAddSessionItemsBody
exclusiveWhen true, each item is first detached from every other session it belongs to (same item type) before being attached here - i.e. 'move' instead of 'add'. Without it, a task another started session holds is refused with 409 task_in_active_session (a task works in one session at a time): pass true to move it here, after asking the person.
CancelSessionBody
discardPlanSnapshot the drafted plan into Document history, then clear the session description
DraftSessionPlanBody
modeBuild mode to lock in before dispatch (cloud | codeflow_local)
codeflowInstanceIdTarget local Productmind agent instance id (required when mode = codeflow_local)
RetrySessionRunBody
modeBuild mode to lock in before dispatch (cloud | codeflow_local)
codeflowInstanceIdTarget local Productmind agent instance id (required when mode = codeflow_local)
ResumeSessionBody
modeBuild mode to lock in before dispatch (cloud | codeflow_local)
codeflowInstanceIdTarget local Productmind agent instance id (required when mode = codeflow_local)
RequestSessionTranscriptUploadBody
clientThe coding agent you are running in: claude_code (Claude Code), codex (Codex CLI or IDE extension) or copilot (GitHub Copilot CLI, or Copilot Chat agent mode in VS Code). Narrows where the command looks for your transcript; omit it when unsure and every supported location is searched.
shellThe shell your command tool runs: posix (the default: macOS, Linux, WSL, and the Git Bash Claude Code uses on Windows) or powershell (Windows PowerShell or PowerShell 7).
SessionInteractionOption
valuelabeldescriptionrecommendedSet true on the option you recommend (put it first). The app labels it Recommended and picks it in advance, so the user can accept it with one click. Omit it when you have no recommendation; on a single-choice question flag at most one.
RespondToSessionInteractionBody
selectedcustomTextdecisionreadOnlynoteQueuePrReviewChangeBody
contentYour reply in the thread, plain text (no markdown): one to three short lines saying what you will change. It is posted as your answer, and the thread joins the review queue that Apply code revisions sends.
RespondToCiChangeBody
decisionapprove commits every open entry onto its branch; decline leaves the pull request without them
AttachSessionRepoBody
workforceConnectorIdexternalRepoIdrepoNamerepoUrlproviderbaseBranchworkingBranchreadOnlytrue attaches the repository for context only: read-only in this session, never committed, pushed or opened as a pull request on. Leave it out to inherit: a repository is read-only anyway when an application on it is.
AttachSessionApplicationsBody
applicationIdsApplications in this workspace whose repositories the session should work in. Each must have a repository connected; a repository the session already has is skipped. Each one also runs in the session's sandbox, even when another sandbox lists it.
readOnlytrue attaches these repositories for context only: read-only in this session, never committed, pushed or opened as a pull request on. Leave it out to inherit each application's own Read-only setting.
UpdateSessionRepoBody
baseBranchThe session's own read-only mark: true makes the repository read-only in this session, false makes it writable again, null clears the mark. A repository an application on it makes read-only stays read-only: false is refused there.
UpdateSessionPullRequestDescriptionBody
urlHTML URL of the pull request whose description to write
bodyThe new description (markdown). An empty string clears it.
baseBodyThe description the edit started from, as it was read. When the pull request's description has changed since, nothing is written and the answer is a 409 carrying the current description.
SessionApprovePullRequestBody
urlHTML URL of the pull request to act on
forceMerge even when the pull request's checks are not green. Skips the CI gate but NOT the provider's own branch protection — a merge the provider refuses still fails.
headSha^[0-9a-f]{7,40}$The head commit SHA of the pull request as you reviewed it. When given, the merge is refused (409) if the pull request's head has moved since, and the provider is asked to merge exactly that commit. Required for Productmind support access.
SessionPullRequestCheckRerunBody
urlHTML URL of the pull request to act on
externalId^\d{1,20}$ · requiredThe check to re-run: the externalId of a row from the pull request's checks list (a GitHub check-run id or a GitLab pipeline id).
SessionPullRequestMergeTargetBody
urlHTML URL of the pull request to act on
resolveConflictstrue: if merging the target branch in conflicts, the agent resolves the conflicts in a new turn and pushes a merge commit. false: a conflict ends the turn with nothing pushed and the conflicted files listed.
OpenPullRequestBody
repoNameRepository name (matches a session repo)
titlePull request title
bodyPull request body (markdown)
DeploySessionBody
commitShaenvironment^[a-z](?:-?[a-z0-9])…The environment's key: development, production, or one of the workspace's own environments.
deployedBybuildUrlnotesCreateSessionDeploymentBody
environment^[a-z](?:-?[a-z0-9])… · requiredThe environment's key: development, production, or one of the workspace's own environments.
commitShatagNamedeployedBydeployedAtstatusbuildUrlpullRequestUrlnotesUpdateSessionDeploymentBody
statusStartSessionPreviewBody
sandboxIdttlSecondsidleTimeoutSecondsmodecodeflowInstanceIdCreateTagBody
entityentityIdnameThe tag's name. Saved lowercase-with-hyphens: lowercased, every run of characters that is not a letter or digit becomes one hyphen, trimmed of hyphens and capped at 40 characters (Billing & Payments is saved as billing-payments). A name the entity already carries in any spelling (Billing for billing) reuses that tag. A name with no letter or digit is refused.
TagGenerateBody
entityentityIdsTagSuggestionsAcceptBody
entityentityIdsincludeNewNamesTagSuggestionNameBody
entitynameThe proposed name exactly as listed by the new-names read: matched as stored, so a suggestion staged before the lowercase-with-hyphens rule stays reviewable.
AgentContextDTO
idworkspaceIdtypetitlesortOrderisStandardisCustomizedcreatedAtupdatedAtAgentDefinitionDTO
idworkspaceIdslugnameinstructionsversionenabledtoolstoolsRequiringApprovaloriginasyncEligiblecreatedAtupdatedAtAgentToolCatalogEntry
namesourcemutatingasyncEligibledescriptionWorkspaceModelClassesDTO
platformRuntimecanEditplatformCodexModelIdsWorkspaceModelClassModelDTO
modelIddisplayNameisActiveWorkspaceModelClassTierDTO
modelClassWorkspaceModelClassCellDTO
runtimecustomizedWorkspaceModelClassEntryDTO
modelIdrankmaxEffortCustomizedavailableAutomationDTO
idworkspaceIdnameenabledrunAsUserIdconsecutiveFailurescreatedAtupdatedAtAutomationRunSummary
idstatuscreatedAtConditionVerdict
matchedWhether the rule would fire. Un-checkable leaves count as a pass
unevaluableConditionLeafVerdict
indexDepth-first position among the rule's condition leaves
fieldoperatoroutcomeactualWhat the entity actually held
reasonWhy the condition could not be answered
AutomationCheckResult
entityTypeAutomationRunDTO
idworkspaceIdagentSlugtriggerTypestatusrunAsUserIddepthoutputTruncatedcreatedAtupdatedAtCommentDTO
identityentityIdcontentkindparentCommentIdresolvedAtauthorKindaddressedTocreatedAtupdatedAtCommentAgentTurn
statusmoderequestedByUserIdrequestedAtstartedAtfinishedAtrequestIdchatIdsessionIdsessionDisplayIdproposalIdansweredThroughAterrorCommentExternalAuthorDTO
sourceshareIdviewerIdnameemailinteractionIdDocumentPassageEditDTO
The edit as the thread records it: Undo puts before back.
blocksThe highlighted blocks as they read now, in markdown.
DecisionDTO
identityentityIddecisionTypeoutcomedecidedAtisSupersededisCurrenttitletoStatusrationaleconditionalNotesdecidedByUserIddecidedByNamedecidedByUserNamesupersedesIdcreatedAtupdatedAtPolymorphicAttachmentDTO
idfileNameurlcontentTypefileSizeentityentityIdcreatedAtupdatedAtoriginvisibilityWorkspaceDTO
idnamedescriptionstatusThe workspace's logo image URL, or null when it has none
organizationIdYour own role on this workspace: owner, admin or member
OnboardingSteps
introductionverificationapplicationplanproductContexttaskscodecodingAgentsslackpaymentreviewOnboardingProfile
buildIntentteamSizeFeatureListItemDTO
iddisplayIdworkspaceIdkindtitleprioritystatussortOrderparentFeatureIdlayerdescriptionrequestingAccountIdrequestingContactIdownerIdestimatedEfforttargetDateexternalSourceexternalIdexternalSourceUrldescriptionEnrichmentStatussubFeatureCountrestrictedTrue on a placeholder for a private feature the caller cannot see: its title is withheld and nothing of its content is included.
createdAtupdatedAttagssuggestedTagCountspecificationStateWhere the feature's specification stands: 'queued' (a deep dive is dispatched, no agent has picked it up yet), 'scanning' (an agent is reading the code now), 'awaiting_review' (a proposed rewrite is waiting on a human), 'pending' (created by the repository map pass, no specification written yet). Absent when there is nothing to say. Populated by the list endpoint only.
specificationProposalIdId of the in-flight FeatureDescriptionProposal behind specificationState. Present only while that state is 'queued' or 'scanning' — it is what a Stop affordance cancels. Populated by the list endpoint only.
specificationSourceWhich flow produced the in-flight proposal behind specificationState: a scan of the code ('refresh', 'deep_dive', 'composition', 'discovery') or a revision from review comments ('review'). Present only while specificationState is 'queued' or 'scanning'. Populated by the list endpoint only.
specificationErrorWhy the last run failed. Present only when specificationState is 'failed' — it is what lets a list row say what broke instead of only that something did. Populated by the list endpoint only.
specificationStartedByIdId of the user who started the in-flight scan. Present only while specificationState is 'queued' or 'scanning', and only when the scan recorded who started it. Populated by the list endpoint only.
The user who started the in-flight scan (name and avatar). Present only while specificationState is 'queued' or 'scanning'. Populated by the list endpoint only.
Where the in-flight scan runs: its scan session's build mode, status and local agent name. Present only while specificationState is 'queued' or 'scanning'. Populated by the list endpoint only.
The scan of the code that finished last and where it ran: its proposal id and its scan session's build mode, status and local agent name. Present for 24 hours after the run ended, while the latest proposal is that scan with its result waiting for review or applied; absent after a scan that failed, was stopped or was dismissed, after a later change that is not a scan, and while a new scan is queued or running. Populated by the list endpoint only.
deepScanStatusWhether a scan of the code has written this feature's specification: 'scanned' (it has), 'awaiting_review' (a scan's draft is waiting for review, even when an earlier scan already wrote one), or 'not_scanned' (no scan has written one yet). The value the deepScan filter matches. Populated by the list endpoint only.
LinkedApplicationDTO
idnamecodePathCountentryPointCountLinkedServiceDTO
idnamekindlayersourceFeatureViewerAccess
isPrivateTrue when the feature is private: it, or a feature above it, lists the teams that may see it.
canEditThe caller may edit the feature: fields, document, status, links, business rules, and reviewing proposals.
canDeleteThe caller may delete the feature.
canAddSubFeatureThe caller may create a sub-feature under it.
canManageAccessThe caller may change which teams can see and edit the feature.
FeatureBreadcrumbDTO
iddisplayIdtitlekindparentFeatureIdrestrictedFeatureScanRunnerDTO
statusFeatureCompletedScanDTO
proposalIdFeatureDTO
iddisplayIdworkspaceIdkindtitleprioritystatussortOrderparentFeatureIdlayerdescriptionrequestingAccountIdrequestingContactIdownerIdestimatedEfforttargetDateexternalSourceexternalIdexternalSourceUrldescriptionEnrichmentStatussubFeatureCountrestrictedTrue on a placeholder for a private feature the caller cannot see: its title is withheld and nothing of its content is included.
createdAtupdatedAtdescriptionLexicaldescriptionUpdatedAtdescriptionVersionmentionedUserIdstagssuggestedTagCountAccountDTO
idorganizationIdnameaccountTypewebsitenotessourceTypesourceIdexternalSourceWhich external system this record was imported from.
externalIdThe CRM's own id for this record — people/c123…, 701.
externalSourceUrlDeep link back to the record in the CRM.
isStrategicestimatedArrlifecycleStatetagssuggestedTagCountcontactCountcreatedAtupdatedAtContactDTO
idaccountIdnameisPrimaryaccountNameemailtitlephonesourceTypesourceIdexternalSourceWhich external system this record was imported from.
externalIdThe CRM's own id for this record — people/c123…, 701.
externalSourceUrlDeep link back to the record in the CRM.
lastContactedAtcreatedAtupdatedAtFeatureDeleteImpactDTO
subFeatureCountSub-features deleted with the feature, at any depth. The feature itself is not counted.
taskCountDistinct tasks deleted with the feature and its sub-features, subtasks included.
FeatureRequestDTO
iddisplayIdworkspaceIdtitledescriptionprioritystatussourcesortOrderinterestCounttagssuggestedTagCountestimatedEffortassigneeIdcreatedAtupdatedAtFeatureRequestInterestDTO
idsourcehasSourceUpdatessourceEntityIdsourceItemIdsourceTimestampAtImportsourceNamesourceDescriptionsourceNavigationUrldescriptionsentimentrequestingAccountIdrequestingContactIdpositionconditionalNotesstakeholderRolestakeholderUserIdstakeholderNameweightcreatedAtFeatureRequestLinkedEntity
iddisplayIdtitleentityTyperelationshipTyperestrictedkindlayerexternalSourcetaskTypesourceBugReportSessionSummaryDTO
idnamecategorystatusdisplayIdFeatureRequestRelationshipDTO
idtargetTypetargetIdrelationshipTypedisplayIdtitletargetKindrestrictedFeatureRequestAccountInterestDTO
countlatestAtdominantSentimentFeatureLogDTO
idcontentcreatedAtupdatedAtFeatureAccessDTO
featureIdrestrictedTrue when the feature is private, through its own list or a list on a feature above it.
The nearest feature above this one whose list governs it, when this feature has no list of its own.
This feature's own list. Empty when it has none.
The list of the feature named in inheritedFrom. Empty when the feature has a list of its own or is open.
The sub-features below this feature that do not simply follow its list. A change to this feature's list can make the changeable ones follow it (subFeatures: follow).
FeatureAccessSource
iddisplayIdtitlekindFeatureAccessGrantDTO
teamIdlevelThe team, or null when it has been deleted. A grant to a deleted team keeps the feature private until it is removed.
FeatureAccessTeamRef
idnamememberCountGenerateTasksResponse
GeneratedTask
typetitleprioritydescriptiontagsLayer tags for the task (e.g. backend, frontend, agent, test, docs, infra).
FeatureSearchResult
scorematchTypeWhich search path produced this hit: fulltext (PostgreSQL tsvector / SQLite LIKE — exact keyword), or semantic from the KB fallback (LanceDB — semantic-only for entity titles; score is absolute semantic similarity, comparable across queries; sub-floor hits are filtered server-side, so an empty list means no match).
FeatureVersionDTO
idfeatureIdversionNumbertitleauthorIdsdescriptionMarkdownonBehalfOfUserIdThe person an agent author of this version worked for: who started its session, or who the integration acted for. Absent when a person saved it themselves.
The person in onBehalfOfUserId. Absent when they no longer have an account.
summarycreatedAtFeatureVersionDetailDTO
idfeatureIdversionNumbertitleauthorIdsdescriptionMarkdownonBehalfOfUserIdThe person an agent author of this version worked for: who started its session, or who the integration acted for. Absent when a person saved it themselves.
The person in onBehalfOfUserId. Absent when they no longer have an account.
summarycreatedAtdescriptionLexicalLinkedDocumentDTO
idtitlesourceTypestatussortOrderWhere a repository-synced document comes from; null for every other document
LinkedCodeSurfaceEcho
applicationIdnamecodePathsentryPointsmanualCodePathsmanualEntryPointscodePathCountentryPointCountFeatureCodeSurfaceDTO
applicationIdnamecodePathsentryPointsmanualCodePathsmanualEntryPointshiddenCodePathsmanualSystemComponentIdshiddenSystemComponentIdsArchitectureServiceDTO
idnamekindlayerFeatureDependencyLinkDTO
directionFeatureBusinessRuleDTO
idfeatureIdbatchIdprovenanceruleKeyruleNumbersortOrdertitleruleTextenforcedAtsourceDecisionIdsverificationStatuscurrentcreatedAtupdatedAtfeatureDisplayIdfeatureTitlefeatureKindSubmitBusinessRulesResult
batchIdwrittenbatchTotalpromotedskippedForLimitFeatureDescriptionProposalDTO
idworkspaceIdfeatureIdsourcekindstatusdraftStalecreatedAtupdatedAtProposalUsageTotalsDTO
inputTokensoutputTokenscacheReadTokenscacheWriteTokensFeatureReviewRevisionDTO
FeatureRefreshStatusDTO
statusnewCommitCountcappedFeatureDescriptionProposalReceipt
idfeatureIdsourcekindstatuscharactersguidanceparksAutoApplyCodesFeatureRequestSearchResult
idFeature request UUID
titlescoreAbsolute semantic similarity (cosine-derived). Higher is better and scores ARE comparable across queries. Results below the server-side noise floor are filtered out.
matchTypeAlways "semantic" for KB-backed hits — entity title search runs semantic-only. Sub-floor hits are dropped server-side, so an empty list means no match.
TaskParentDTO
entityTypeidtitledisplayIdkindlayerexternalSourceownerIdapplicationTyperestrictedFeatureRequestSourceSummaryDTO
sourcesourceEntityIdsourceItemIdsourceNamenavigationUrlisSourceAvailableimportedAtsourceDescriptionFeatureRequestLogDTO
idcontentcreatedAtupdatedAtFeatureRequestDescriptionProposalDTO
idworkspaceIdfeatureRequestIdsourcestatusWhy the run declined to write anything, when it settled failed with sufficientContext false and the AI named a reason. Null otherwise.
createdAtupdatedAtBugReportDTO
iddisplayIdworkspaceIdtitlepriorityseveritystatussortOrderdescriptionreporterCountassigneeIdtagssuggestedTagCountcreatedAtupdatedAtBugReportReporterDTO
idtypereporterIdnameemailavatarUrlinitialsdetailscreatedAtBugReportRelationshipDTO
idtargetTypetargetIdrelationshipTypedisplayIdtitletargetKindBugReportConvertedEntityDTO
entityTypeidtitledisplayIdstatusrestrictedBugReportLinkedEntity
idtitleentityTyperelationshipTypedisplayIdInterestDTO
identityentityIdsourcehasSourceUpdatessourceEntityIdsourceItemIdsourceTimestampAtImportsourceNamesourceDescriptionsourceNavigationUrldescriptionsentimentrequestingAccountIdrequestingContactIdstakeholderRolestakeholderUserIdstakeholderNamecreatedAtConvertBugReportResponse
entityTypecreatedBugReportDescriptionProposalDTO
idworkspaceIdbugReportIdsourcestatusWhy the run declined to write anything, when it settled failed with sufficientContext false and the AI named a reason. Null otherwise.
createdAtupdatedAtStartTaskWorkResponse
SessionDTO
idworkspaceIdnamecategorystatusautonomouspullRequestUrlschatIddisplayIdreleaseIddescriptiondescriptionLexicaldescriptionUpdatedAtdescriptionVersionmentionedUserIdsassigneeIdbranchNametagNamerepositoryUrlhasCommitsinterruptSeqinterruptReasondispatchRunKindretryContinuesBuildretryContinuesRevisionretryContinuesPlanRevisiontransportPinnedphaseclaudeSessionIdplanApprovedAtplanApprovedByIdplanApprovedMatchesDescriptionplanUnchangedSinceApprovalplannedDaterelatedSessionIdcreatedByupdatedByscopeSeqtagssuggestedTagCountcreatedAtupdatedAtSessionScanSubject
kindrepositoriesrescanSessionPullRequestFailure
repoNamesteperrorpullRequestOwedatcodepathsSessionPublishFailureKept
head^(?:[0-9a-f]{40}|[0-… · requiredcommitsstoredSessionCreditStop
atbalanceCentsthresholdCentscurrencySessionBillingRefusal
codeatSessionGitAccessFailure
codeatSessionRunFailure
codestagetitlemessageoccurredAtSessionTargetMerge
pullRequestUrlrepoNametargetBranchresolveConflictsrequestedAtSessionPlanQuestion
statecommentIdsstartedAttargetSessionSpendingPause
statepausedAtSessionCiAutofixSummary
enabledattemptsmaxAttemptsSessionRunTimeSummary
agentMsturnCountcompleteSessionInteractionDTO
idsessionIdphasekindquestionmultiSelectstatusclaudeSessionIdrespondedByIdrespondedAtcreatedAtupdatedAtawaitsViewerSessionInteractionQuestion
keyquestionmultiSelectSessionCiChangeOptions
versioncontentsOmittedSessionCiChangeEntry
repoNamesessionGitRepositoryIdownerrepoworkingBranchreferenceRefreferenceShaoriginalHeadShasanitizedHeadShareportedAtstatusreportedHeadShasSessionCiChangeFile
pathafterSizeBytesSessionRepositoryRequest
repoNameapplicationIdsapplicationNamesreasonsuggestedReadOnlylockedReadOnlydepartingIdSessionItemDTO
idsessionIditemTypeitemIditemNameitemStatusdisplayIdtaskTypefeatureIdfeatureNamefeatureRestrictedaddedBynotescreatedAtSessionDeploymentDTO
idsessionIdenvironmentThe environment's key: development, production, or one of the workspace's own.
statuscommitShatagNamedeployedAtdeployedBybuildUrlpullRequestUrlnotescreatedAtupdatedAtFor an environment that deploys to a cloud account: the newest deploy run of this commit there, as GitHub ran it and Productmind recorded it, beside this row's own status. Null when none is recorded.
DeploymentRunSummaryDTO
deploymentIdkinddeploy: a run of the repository's deploy workflow; destroy: a run of its destroy workflow.
statusWhere the run is: queued (waiting for its turn or a runner), in_progress, or completed.
runNumberrunAttemptSessionGitRepositoryDTO
idworkforceConnectorIdexternalRepoIdrepoNamerepoUrlbaseBranchworkingBranchmodifiedpullRequestUrlreadOnlySessionDepartingRepositoryDTO
idrepositoryIdrepoNamerepoUrlbaseBranchworkingBranchreadOnlystateremovedAtSessionPendingScopeNoticeDTO
idkindsubjectIdSessionPreviewDTO
idsessionIdworkspaceIdstatusmodettlSecondsidleTimeoutSecondsunsavedWorkcreatedAtupdatedAtSessionPreviewUrl
portnameurlSessionPreviewApp
applicationIdnameworkingDirectoryinstallCommanddevCommandSessionPreviewAppAiRepair
configuredInstallCommandThe install command the application held when the repair was offered
configuredDevCommandThe run command the application held when the repair was offered
installCommandThe install command the repair got working
devCommandThe run command the repair got working
savedAtWhen the repaired commands were saved on the application (ISO 8601)
savedByWho saved them: a user id
SessionPreviewAiSetup
admittedWhether the AI-assisted setup turns may run
reasonWhy they will not: the coding check's sentence
CompleteTaskWorkResponse
TaskLogDTO
idcontentcreatedAtupdatedAtTaskDescriptionProposalDTO
idworkspaceIdtaskIdsourcestatusWhy the run declined to write anything, when it settled failed with sufficientContext false and the AI named a reason. Null otherwise.
createdAtupdatedAtCreateContactResult
idaccountIdnameisPrimarymatchedExistingtrue when a contact with this email address was already on file: this is that contact, returned unchanged, and nothing was created. false when this call created the contact.
accountNameemailtitlephonesourceTypesourceIdexternalSourceWhich external system this record was imported from.
externalIdThe CRM's own id for this record — people/c123…, 701.
externalSourceUrlDeep link back to the record in the CRM.
lastContactedAtcreatedAtupdatedAtNoteDTO
idworkspaceIdtitlecontactIdrawContenttagssuggestedTagCountpinnedcreatedAtupdatedAtNotePromotedEntityDTO
entityTypeidtitledisplayIdstatusrestrictedPromoteNoteResponse
entityTypecreatedNotificationDTO
idtypeemailSkippedorganizationIdworkspaceIdsubjectTypesubjectIdreadAtemailSentAtdirectMessageSentAtdirectMessageProviderdirectMessageStatuscreatedAtupdatedAtCustomerInteractionDTO
workspaceIdtypeidcustomerResearchIdcontactIdstatustagssuggestedTagCountrawContentguidelinescheduledAtendsAtmeetingUrlmeetingProvidervoiceEnabledrecordingEnabledtranscriptionEnabledexternalSourceexternalIdexternalSourceUrlagentInterviewLinkcreatedAtupdatedAtInteractionParticipant
emailnameroleCustomerInteractionMessageDTO
contentidtypeisFromAgentinteractionIdcreatedAtupdatedAtChatMessageAttachment
attachmentIdfileNamecontentTypeInteractionSupportRequestRef
iddisplayIdtypestatuscreatedAtCustomerInteractionDescriptionProposalDTO
idworkspaceIdinteractionIdsourcestatusWhy the run declined to write anything, when it settled failed with sufficientContext false and the AI named a reason. Null otherwise.
createdAtupdatedAtInternalMeetingDTO
workspaceIdtitlestatusidagendanotestagssuggestedTagCountscheduledAtendsAtmeetingUrlmeetingProviderrecordingEnabledtranscriptionEnabledexternalSourceexternalIdexternalSourceUrlcreatedAtupdatedAtInternalMeetingParticipant
emailnameuserIdroleInternalMeetingSummary
titlesummarykeyPointsdecisionsInternalMeetingRelationshipDTO
idtargetTypetargetIdrelationshipTypedisplayIdtitletargetKindSupportImportPreviewItem
externalIdisNewImportcontactIdcontactNameaccountIdaccountNameexistingInteractionIdexistingInteractionTypeexistingContactIdexistingContactNameexistingAccountNamedefaultInteractionTypeambiguousContactMatchotherHelpdeskAccountSupportImportResult
createdCountupdatedCountfilteredCounterrorCountSupportImportResultItem
externalIdoutcomeinteractionIdinteractionTypecontactIdaccountIdmessageSupportIngestionConfigDTO
platformstatusestagstagMatchModegroupIdsdefaultInteractionTypeidlastWebhookAtcreatedAtupdatedAtSupportIngestionEventDTO
idplatformoutcomesourceexternalIdmessageinteractionIdcreatedAtSupportRequestListResponse
Pass back as cursor for the next page; null on the last page.
SupportRequestListItemDTO
Decision Table
| Variant | Matching Criteria |
|---|---|
| type = object · tier="full" · requires: id, displayId, sequenceNumber +12 more | |
| type = object · tier="billing" · requires: id, displayId, sequenceNumber +10 more |
tieriddisplayIdsequenceNumbertypestatustitlelastActivityAtcreatedAtunreadSupportRequestFullListItemDTO
tieriddisplayIdsequenceNumbertypestatustitlelastActivityAtcreatedAtunreadSupportPersonDTO
displayNamesideSupportRequestBillingListItemDTO
tieriddisplayIdsequenceNumbertypestatustitlelastActivityAtcreatedAtSupportRequestDTO
Decision Table
| Variant | Matching Criteria |
|---|---|
| type = object · tier="full" · requires: id, displayId, sequenceNumber +26 more | |
| type = object · tier="billing" · requires: id, displayId, sequenceNumber +19 more |
tieriddisplayIdsequenceNumbertypestatusstatusChangedAtorganizationIdtitlelastActivityAtcreatedAtupdatedAtpriorityunreadSupportRequestFullDTO
tieriddisplayIdsequenceNumbertypestatusstatusChangedAtorganizationIdtitlelastActivityAtcreatedAtupdatedAtpriorityunreadSupportRequestSubjectDTO
typeidSupportRequestBillingDTO
currencySupportQuoteDTO
idsupportRequestIdkindstatuscapMinutesrateCentsPerHourmaxAmountCentssubmittedAtexpiresAtSupportRequestAccessSummaryDTO
allowMergeeverIssuedSupportScreenshotDTO
idfileNamecontentTypefileSizesideurlcreatedAtSupportRequestViewerDTO
tierisRequestercanPostMessagecanCancelcanDecideQuotescanRevokeAccessSupportRequestBillingTierDTO
tieriddisplayIdsequenceNumbertypestatusstatusChangedAtorganizationIdtitlelastActivityAtcreatedAtupdatedAtSupportRequestUnreadCountDTO
countawaitingCountSupportRequestMessageListResponse
Pass back as cursor for the next page; null on the last page.
SupportRequestMessageDTO
idsupportRequestIdkindauthorSidecreatedAtcanEditSupportRequestScopePreviewDTO
hiddenCountSupportRequestScopeSessionDTO
idnamestatusSupportRequestSharedFeatureDTO
sourceremovableSupportRequestSharedTaskDTO
sourceremovableSupportRequestSharedBugReportDTO
sourceremovableSupportRequestSharedUnavailableDTO
typeidsourceremovableSupportRequestSharedEntitiesDTO
accessScopeVersioncanSharehiddenCountSupportRequestEligibilityDTO
typeeligiblebillablecurrencycanAllowMergeSupportRequestChipDTO
iddisplayIdtypestatuscreatedAtSupportRequestChipListResponse
SupportAccessGrantOrgListResponse
Pass back as cursor for the next page; null on the last page.
SupportAccessGrantOrgDTO
idstatuscontentVisiblelevelallowMergeissuedAthardExpiresAtSupportAccessGrantOrgResponse
ProductmindSupportAccessResponse
ProductmindSupportAccessDTO
organizationIdenabledrequiresSsodefaultFromSsoactiveGrantCountopenAccessRequestCountProductmindSupportAccessUpdateResponse
ProductmindSupportAccessUpdateDTO
organizationIdenabledrequiresSsodefaultFromSsoactiveGrantCountopenAccessRequestCountgrantsRevokedrequestsCancelledaccessEndedsweepCompleteChatMessageDTO
idcontentisFromAgentsessionIdsourcekindpendingInterruptIdcreatedAtupdatedAtChatReference
entityTypeentityId^([0-9a-fA-F]{8}-[0-… · requiredtitledisplayIdkindtaskTypesessionCategoryCustomerResearchDTO
workspaceIdtitleobjectivesquestionsiddescriptionguidelineinterviewWelcomeMessagewidgetWelcomeMessageinterviewerAgentIdwidgetKeystatuscreatedAtupdatedAtsummaryGeneratedAtcompletedInteractionCountAtGenerationinteractionCountcompletedInteractionCountisSummaryStaleFeatureRequestSummaryDTO
iddisplayIdtitleprioritystatusinterestCountTagVocabularyEntryDTO
nameTag name
countHow many entities of the requested type carry the tag
KnowledgeBaseFollowOnDTO
overviewRefreshTrue when a source indexed now would redraft the Overview: it was never generated, or generated and never edited
starterFeaturesTrue when the workspace has no features yet: indexing the Overview drafts starter features with their specifications
A measured estimate of what the follow-on work draws; 0 when nothing would run
KnowledgeBaseDocumentListItemDTO
idworkspaceIdorgIdtitlesourceTypestatusisStaleisSystemisWiredavailableToAgentsdisplayIds3KeymimeTypetagssuggestedTagCountdescriptionchunkCountchunkingStrategysortOrdercreatedAtupdatedAtKnowledgeBaseDocumentMetadata
providerexternalIdexternalUrloriginalFileNamegenerationInsufficientgenerationNotegeneratedContentHashrefusalCodeKnowledgeBaseRepoSyncMetadata
syncIdThe Knowledge Base sync that owns this document
applicationIdThe Application whose repository the document is synced from
kindroot: the sync's root document; folder: a folder; file: a Markdown file
pathThe file or folder path in the repository; empty for the root document
providerThe git provider: github, gitlab or bitbucket
repoNameThe repository's full name, such as owner/repo
branchThe branch the document is synced from
webUrlThe file or folder on the provider's website, at the synced branch
commitShaThe commit the document's content was last synced from
KnowledgeBaseDocumentDTO
idworkspaceIdorgIdtitlesourceTypestatusisStaleisSystemisWiredavailableToAgentsdisplayIds3KeymimeTypetagssuggestedTagCountdescriptionchunkCountchunkingStrategysortOrdercreatedAtupdatedAtKbDocumentLinkDTO
identityTypeentityIdKnowledgeBaseDocumentProposalDTO
idworkspaceIdkindsourcestatusdraftStalecreatedAtupdatedAtKnowledgeBaseDocumentProposalSummaryDTO
idkindsourcestatuscreatedAtKnowledgeBaseSearchResponseDTO
querymodetotalusageEventIdKnowledgeBaseRagResponseDTO
answerquerymodetotalResultsusageEventIdKnowledgeBaseIndexingDeferral
deferredreasonWhy indexing waits: insufficient_credit, payg_activation_required, spending_paused or a usage limit
messageThe sentence that says why, and what fixes it
IngestConnectorItemsResponseDTO
executionIddocumentIdsPresent when the imported documents will be stored but not indexed until the organization can pay
KnowledgeBaseDocumentVersionDTO
iddocumentIdversionNumbertitleauthorIdscontentMarkdownonBehalfOfUserIdThe person an agent author of this version worked for: who started its session, or who the integration acted for. Absent when a person saved it themselves.
The person in onBehalfOfUserId. Absent when they no longer have an account.
summarycreatedAtKnowledgeBaseDocumentVersionDetailDTO
iddocumentIdversionNumbertitleauthorIdscontentMarkdownonBehalfOfUserIdThe person an agent author of this version worked for: who started its session, or who the integration acted for. Absent when a person saved it themselves.
The person in onBehalfOfUserId. Absent when they no longer have an account.
summarycreatedAtcontentLexicalChatDTO
idworkspaceIdstatustitlemodeautonomyLevelvisibilityactiveSessionIdsessionIdsagentIdexternalThreadRefcreatedByIdcreatedByNamemessageCountcreatedAtupdatedAtChatSessionSummary
idnamecategorystatusdisplayIdUserDTO
idfirstNamelastNamedisplayNameinitialsemailmobilejobTitleprofilePictureUrldisabledcreatedAtupdatedAttimezonetimeformathasPasswordEntityWorkspaceDTO
workspaceId^([0-9a-fA-F]{8}-[0-… · requiredworkspaceNameUserSessionDTO
idnamesourcecurrentfirstPartyFeatureBoardDTO
unlinkedRequestCountFeatureBoardItemDTO
featureIdparentFeatureIdLineageGraphDTO
rootEntityIdrootEntityTypeLineageNodeDTO
identityTypestagetitlestatusdisplayIddescriptionprioritytypecustomerResearchIdcreatedAtupdatedAtLineageInterestSummary
sourcesourceNamedescriptionsentimentpositionconditionalNotesweightstakeholderRolestakeholderNamecontactNameaccountNameaccountIsStrategicLineageEdgeDTO
sourceIdsourceTypetargetIdtargetTyperelationshipDashboardStatsDTO
DashboardNodeCountsDTO
interviewsAgenticVoiceinterviewsAgenticEmailfeedbackInboundinterviewsManualNotessupportCasesinterestCustomerinterestStakeholderinternalMeetingsnotesrequestsNewrequestsInReviewrequestsPromotedrequestsRejectedfeaturesActivefeaturesGeneratedPendingtasksTodotasksInProgresstasksInReviewtasksDonebugReportsTodobugReportsInProgressbugReportsInReviewbugReportsDonesessionsActivesessionsPlanningsessionsNeedsYousessionsBuildingsessionsCompletedreleasesShippedapplicationsTotalapplicationsPlatformHostedsessionsTotalsessionsRunDashboardKpisDTO
activeConversationsrequestsValidated7dreleasesShipped7dDashboardWindowsDTO
windowDayscustomerFeedbackCreatedinterestCustomerCreatedinterestStakeholderCreatedinternalMeetingsCreatednotesCreatedrequestsCreatedbugReportsCreatedtasksCreatedsessionsStartedDashboardEventDTO
idkindtitleoccurredAtDashboardUpdatesDTO
DashboardSessionMetricDTO
totalDashboardSessionItemDTO
idtitlestatusupdatedAtcategoryDashboardTaskItemDTO
idtitlestatusupdatedAttypeDashboardBugItemDTO
idtitlestatusupdatedAtseverityDashboardRequestMetricDTO
totalDashboardRequestItemDTO
idtitlestatusupdatedAtsourceDashboardFeedbackMetricDTO
totalDashboardFeedbackItemDTO
idtitlestatusupdatedAttypeDashboardTrendsDTO
bucketDaysbucketCountstartDashboardAttentionDTO
DashboardItemDTO
idtitlestatusupdatedAtWorkCountsDTO
sessionsAwaitingMetasksOpenMinebugReportsOpenMineapplicationsScanningsessionsInProgressfeaturesScanningfeaturesAwaitingReviewfeaturesDeepScanningkbProposalsAwaitingReviewautomationRunsInFlightautomationRunsFailedRecentfeatureRequestsOpenreleasesBeingPreparedreleasesReadyToMergemarketingReactionsEntitySearchHit
entityTypeentityIdtitlematchTypescoreDisplayIdLookup
entityTypeentityIddisplayIdTeamDTO
idnameorganizationIdmemberCountTeamRoleDTO
keyWhat role takes on add-team-member and update-team-member.
nameAssignableUserDTO
idTheir user id, for assignees (entityType user) and add-team-member.
nameTheir name, or their email when they have not set one.
emailNotificationChannelSubscriptionDTO
idproviderconnectorInstanceIdchannelIdchannelNameeventsisActivecreatedBylastErrorCodelastErrorAtcreatedAtupdatedAtNotificationChannelProviderStatusDTO
providerconnectedmissingScopesdirectMessageCapableconnectorInstanceIdNotificationChannelTargetDTO
idnameisPrivatenumMembersReleaseDTO
idworkspaceIdnamestatusdisplayIddescriptionprepareStatusreleasedAtcreatedByIdtagssuggestedTagCountcanShipcreatedAtupdatedAtReleaseApplicationDTO
idreleaseIdapplicationIdcommitCountcommitsTruncatedcommitCountCappedpullRequestAdoptedstatuspositionapplicationNamecreatedAtupdatedAtReleaseCommitDTO
idreleaseApplicationIdshatitlepositionPullRequestChecksResultDTO
overallStatePullRequestCheckDTO
sourceexternalIdnamestatusrerunSupportedPullRequestCiExpectationDTO
expectedrequiredChecksReleaseDeploymentDTO
idreleaseIdenvironmentThe environment's key: development, production, or one of the workspace's own.
statuscommitShapullRequestUrltagNamedeployedByIddeployedAtbuildUrlnotescreatedAtupdatedAtFor an environment that deploys to a cloud account: the newest deploy run of this commit there, as GitHub ran it and Productmind recorded it, beside this row's own status. Null when none is recorded.
ReleaseProgress
totalTaskscompletedTasksinProgressTaskspendingTaskspercentCompleteReleaseReadinessDTO
ReleaseReadinessRowDTO
releaseIdnameapplicationsTotalapplicationsReadyReleaseShippedWorkDTO
itemCountcommitsTruncatedReleaseShippedGroupDTO
ReleaseCuratedItemDTO
relationshipIditemTypeitemIdReleaseItemProvenanceDTO
itemTypeitemIdderivedReleaseShippedItemDTO
idreleaseIditemTypeitemIdderivedcreatedAtupdatedAtReleaseItemDTO
idreleaseIdfeatureIdpositionfeatureTitlefeatureStatusfeatureDescriptionnotesderivedcreatedAtupdatedAtReleaseSessionLineageDTO
ReleaseLineageFeatureDTO
idtitlestatusdisplayIddescriptionrestrictedReleaseLineageSessionDTO
idnamestatuscategoryReleaseLineageTaskDTO
idtitlestatusdisplayIdtypeReleaseArtifactDTO
idreleaseIdtypestatuscontentfileKeygeneratedAtupdatedByIdcreatedAtupdatedAtMergeDeploymentChecksDTO
overallStatescopecommit: every check on the commit (a merge row); run: the jobs of the deploy run's attempt (a deploy_run row).
liveUpdatesevents: the provider's check events refresh the list (deployment:checks:updated); poll: they do not reach Productmind (GitLab, Bitbucket, or a GitHub commit status still running), so read again while a check runs.
A sentence said in place of the jobs (an older attempt's jobs are on the newer attempt's row), or null.
applicationIdThe application the merge deployed.
commitShaThe merge commit the jobs ran on, as the provider recorded it: 40 hexadecimal characters, or a short SHA where the provider records one (Bitbucket).
The environment the merge deployed to.
PullRequestFilesResultDTO
truncatedtotalChangedFilesPullRequestFileDTO
filenamestatusadditionsdeletionschangesshaPullRequestFileRisk
levelreasonsPullRequestDescriptionDTO
urlThe pull request's URL
bodyThe description as the git provider holds it now (markdown). Empty when the pull request has none.
When a person last saved this description in Productmind. While set, the coding agent's summaries and the release notes no longer replace it. Null when nobody has, and for a pull request linked by hand.
Who last saved this description in Productmind, or null when nobody has.
CloseReleaseResult
releaseIdPull requests that are now closed, including one the provider had already closed.
Pull requests the provider would not close, each with its own sentence. They stay open.
ReleasePullRequestOutcome
releaseApplicationIdThe release-application row the pull request belongs to.
applicationNameThe application's name, or its repository's when it has none: what the result names it by.
The pull request the outcome is about: the one closed, re-opened, opened in its place, or the one that failed.
The provider's own sentence: why a pull request could not be closed or re-opened.
ReopenReleaseResult
releaseIdPull requests that are open again, with the same number.
Pull requests that could not be re-opened, each with the reason, and the new pull request opened in its place from the same branches.
Applications left without an open pull request, each with the reason.
ReleasePullRequestReplacement
releaseApplicationIdThe release-application row the pull request belongs to.
applicationNameThe application's name, or its repository's when it has none: what the result names it by.
The pull request the outcome is about: the one closed, re-opened, opened in its place, or the one that failed.
The closed pull request that could not be re-opened. It stays closed.
The provider's own sentence: why a pull request could not be closed or re-opened.
adoptedTrue when the pull request in its place was already open for the same branches, and the release took it over rather than opening one.
ReleaseNotesProposalDTO
idworkspaceIdreleaseIdsourcestatuscreatedAtupdatedAtShipReleaseResult
releaseIdmergedRelease-application IDs that merged.
Applications the provider refused, with its own message.
skippedRelease-application IDs with nothing to ship (0 commits ahead). Empty when one application is released.
remainingRelease-application IDs still to merge after this call: the applications not released yet, and any that failed.
releasedTrue when no application is left to merge: the release is now released.
MarketingReactionsDTO
availablelikesdislikesMarketingReactionsNotesDTO
notesIdheadlinemarketingReleaseIdThe Marketing release's number (MKT-n).
titleThe Marketing release's title.
likesdislikesMarketingReleaseSummaryDTO
idtitlestatusEvery public release notes document, newest first: what the list's Public release notes row reads.
Every blog post, newest first: what the list's Blog posts row reads.
What the list's Customer emails row reads.
Readers' likes and dislikes on its public notes in What's new, all time, summed over its notes. Null outside Productmind's own workspace, the only one whose notes reach What's new.
createdAtupdatedAtMarketingReleaseSourceCountsDTO
releasesfeaturesfeatureRequestscustomerInteractionsMarketingNotesSummaryDTO
idThe notes document's id.
headlineThe headline of its text; empty before its first draft.
stateIts run in flight, else its newest; null when it never had one.
When its first version was published: what dates it in What's new. Null before.
Readers' likes and dislikes in What's new; null outside Productmind's own workspace.
createdAtupdatedAtMarketingReactionCountsDTO
likesdislikesMarketingReleaseBlogPostStateDTO
idtitleThe post's title, from its text; empty before its first draft.
featureTitlesThe features the post is about that the reader may see: what names a post with no title yet.
stateThe post's run in flight, else its newest; null when it never had one.
MarketingReleaseDetailDTO
Attached sources, grouped by type then newest first.
canDeleteNothing of it was ever published, approved or emailed: it can be deleted. Archive it otherwise.
Its public release notes documents, newest first.
Its blog posts, newest first.
Its media: how many files are ready, and from which campaigns.
MarketingReleaseCoreDTO
idtitlestatuscreatedAtupdatedAtMarketingReleaseSourceDTO
typeidREL-4, FEAT-12, PFR-9; null for customer feedback.
titleThe source's own status, as its own list words it; null for customer feedback.
attachedAtA release's status: publishing waits until every attached release is Released or Deployed.
Customer feedback: the kind of conversation.
Customer feedback: when the conversation happened.
MarketingReadinessDTO
sourceCountSources attached (releases, features, feature requests, customer feedback).
customerFacingFeatureCountFeatures (never private ones) the sources cover: attached ones and those attached releases shipped.
internalOnlyEvery source is a release, and none of their work is under a customer-facing feature.
With no release attached, nothing waits.
MarketingOpenReleaseDTO
idnamestatusBlogPostSummaryDTO
idmarketingReleaseIdtitleThe post's title, or "" before its first draft.
statefeatureTitlesThe titles of the features in scope, in scope order.
passageCountThe post's run in flight, else its newest one.
approvedVersionCountThe post's place on the blog: null until its first publish (Productmind's own workspace only).
createdAtupdatedAtBlogPostScope
featureIdsThe notes document the passages were picked from.
MarketingGenerationDTO
idstatusacknowledgeInternalOnlyA stable code for a failed run (MARKETING_ERROR_CODE).
What a blog post's run was asked for; null on the notes' runs.
createdAtBlogPostPublicationDTO
liveThe post is on the blog now.
slugpublicationIdThe approved version the blog serves (kept while unpublished).
That version's number; null if it cannot be read.
A newer approved version the blog does not serve yet (Update published moves it there); null when none.
When the version the blog serves last changed.
MarketingReleaseBriefDTO
The customer-facing features the sources cover, in the order of their references. Private features are never listed.
Its public release notes documents, newest first. Take customer wording from a live one first.
Its blog posts, newest first.
The background to write from. Team material: never quote it as if customers had read it.
What was already made for it; no campaign when nothing was rendered for it yet. List the files with list-marketing-release-media.
MarketingBriefFeatureDTO
refIts reference in the sources text (F1, F2, ...): the section headed Feature n (ref Fn) is this feature.
idThe feature's id: read its specification with it.
titleshippedInThe attached releases that shipped it (REL-n); empty when none did.
attachedIt was attached to the Marketing release on its own.
reachedThroughThe attached feature requests or customer feedback that led to it (their titles).
MarketingBriefNotesDTO
idThe notes document's id.
headlineThe headline of text; empty before its first draft.
stateliveA published version is live: text is exactly what customers read. False: text is the working draft.
The live version's number; null when no version is live.
textThe live version's text, in markdown, with its images left out. When no version is live, the working draft: customers have not seen it and it may still change.
truncatedtext was cut to fit: read the whole document for the rest.
MarketingBriefBlogPostDTO
idThe blog post's id: read its text as a Marketing document of kind blog-post.
titleThe post's title, or "" before its first draft.
statefeatureTitlesThe titles of the features the post is about.
MarketingMediaListDTO
Every ready file, a campaign's together.
limitHow many files a Marketing release may hold.
MarketingMediaDTO
idThe file's id.
releaseIdThe id of the Marketing release it belongs to.
kindWhat the file is: clip (a video), loop (a short animated image), poster (a clip's still frame), banner (one still at a channel's exact pixels) or captions (a narrated clip's captions file).
statusready once the file was found in storage; pending while only a place was made for it.
sourceWho put it there: a person's upload, the marketing studio's publish, or a coding session.
fileNameThe file's name: its path inside its campaign's output folder, or as the person named it.
contentTypebytessha256The campaign it was rendered from; null for a person's own upload.
The build of that campaign.
Where the same file also stands on a public host, when its campaign was published there; null otherwise.
The coding session that rendered it, when one did.
createdAtMarketingMediaLinksDTO
A link to each ready file's bytes, by the file's id, shown in place.
expiresAtWhen the links stop working. Read again before then.
MarketingMediaUploadsDTO
One link a file, in the order the files were given.
MarketingMediaUploadDTO
mediaIdThe file's id: pass it to confirm-marketing-media once it is uploaded.
fileNameuploadUrlWhere to PUT the file's bytes. It works once the headers below are sent with it, until expiresAt.
Every header the PUT must carry, exactly: with curl, one -H 'name: value' each, and --data-binary @<file>.
expiresAtWhen the link stops working.
ConfirmMarketingMediaDTO
The files found in storage: they are now listed in the Marketing release's media.
The files that are not there, each with why: not uploaded yet, another size than announced, or an id this Marketing release does not have. Upload it and confirm again.
MarketingMediaDownloadDTO
urlA link to the file's bytes, good for a few minutes.
expiresAtMarketingNotesDetailDTO
notesStateWhat the notes are written from, read-only here.
The run in flight, else the newest one (its failure shows until the next run).
Newest first.
MarketingNotesDTO
artifactIdThe notes document's id: the release-notes room's subject.
headlinesummarybodycontentstatusupdatedAtSend back as expectedUpdatedAt when publishing.
editedAfterGenerationThe text differs from what the last generation wrote, so regenerating loses edits.
Bumped by every save of the collaborative document; null until its first.
When the notes were unpublished: no version is live. Null while a version is live or none exists.
createdAtMarketingPublicationSummaryDTO
idversionpublishedAtMarketingPublishCheckDTO
canPublishpossibleCustomerNamesNames of contacts and accounts behind the release's customer evidence that appear in the text.
containsQuotesnextVersioncontentThe text a publish would copy, read together with expectedUpdatedAt.
MarketingPublicationDTO
idversionpublishedAtmarketingReleaseIdnotesIdThe notes document (or blog post) this is a version of.
contentMarketingReaderFeedbackDTO
availablelikesdislikesopenedHow many people opened the full notes in What's new, each counted once.
MarketingReaderFeedbackConversationDTO
interactionIdsourceThe published version that was live when it started.
statustitleThe reader's opening words, or Feedback before they wrote any.
The reader, as the conversation's contact names them.
startedAtpathWhere the conversation opens in the app, without the workspace prefix.
MarketingReaderReactionsPageDTO
availabletotalpagelimitMarketingReaderReactionDTO
nameemailThe organization they were signed into when they reacted; null when unknown or since deleted.
reactionreactedAtBlogPostScopeOptionsDTO
Each published notes document's newest version, newest publish first; empty when none was published.
BlogPostFeatureOptionDTO
idtitlehasSpecificationFalse when the feature's specification is empty.
customerRequestCountFeature requests promoted into the feature.
kindWhich tier the feature is, as its list row draws it.
A component's plane of the stack.
Where the feature was imported from, if anywhere.
statusThe feature's status.
The feature requests promoted into the feature, newest first.
BlogPostCustomerRequestDTO
idtitlestatussourceinterestCountThe interests on the request: the customers who asked for it.
BlogPostNotesOptionDTO
idheadlineversionpublishedAtcontentCustomerNotesSectionDTO
indexheadingtextBlogPostCreatedDTO
BlogPostDetailDTO
idmarketingReleaseIdtitleThe post's title, or "" before its first draft.
statefeatureTitlesThe titles of the features in scope, in scope order.
passageCountThe post's run in flight, else its newest one.
approvedVersionCountcontentstatuseditedAfterGenerationThe text differs from what the last completed run wrote, so regenerating loses edits.
The features in scope, in scope order.
The guidance the newest run used.
Approved versions, newest first.
What stops Approve, in order.
The post's place on the blog: null until its first publish (Productmind's own workspace only).
createdAtupdatedAtBlogPostScopeFeatureDTO
idtitlehasSpecificationFalse when the feature's specification is empty.
customerRequestCountFeature requests promoted into the feature.
inSourcesFalse when the feature is no longer a customer-facing feature of the sources.
kindWhich tier the feature is, as its list row draws it.
A component's plane of the stack.
Where the feature was imported from, if anywhere.
statusThe feature's status.
The feature requests promoted into the feature, newest first.
BlogPostApproveCheckDTO
canApprovepossibleCustomerNamesNames of contacts and accounts behind the post's customer evidence that appear in the text.
containsQuotesnextVersioncontentThe text an approval would copy, read together with expectedUpdatedAt.
BlogPostPublishResultDTO
The deploy of the blog this asked for; null when it could not even be recorded. The post's change stands either way.
BlogDeployOutcomeDTO
repeatedTrue when a deploy of the same posts was asked for moments ago: it is answered and nothing new was started.
BlogDeployDTO
idreasonrevisionThe live set's revision when the deploy was asked for.
staterequested (no answer yet), dispatched (accepted: a run was started), failed (refused: nothing started) or unknown (the answer was lost: a run may have started; it is never tried again by itself).
requestedAtBlogStatusDTO
revisionThe revision of the posts on the blog as Productmind holds them now.
livePostsCustomerEmailOverviewDTO
The live notes documents, newest publish first: Write drafts picks one (the first by default).
What refuses every send now: no live notes, the drafts' notes taken off since, attached releases not Released, a run in flight.
The run in flight, else the newest one (its failure shows until the next run).
When the newest completed run finished; null until one has.
The notes document the newest completed run was written from.
The notes version the newest completed run was written from.
CustomerEmailNotesOptionDTO
idheadlineversionIts live (newest) published version.
publishedAtCustomerEmailRecipientDTO
contactIdThe recipient's own contact: drafts and sends are keyed on it.
contactIdsEvery contact sharing the address.
recipientKeynameeligibilityWhat refuses this recipient now (CUSTOMER_EMAIL_REFUSAL).
CUSTOMER_EMAIL_WARNING: never blockers.
possibleCustomerNamesNames of OTHER contacts and accounts behind the release's customer evidence that appear in the text.
CustomerEmailItemDTO
keyrequest:<featureRequestId> or fix:<bugReportId>: the highlight key.
kindidThe feature request's or bug report's id.
titledeliveredByWhat delivered it: feature or task titles (a request), the bug itself (a fix).
shippedRefsThe sources it came through: release:<id>, feature:<id>, feature_request:<id>, customer_interaction:<id>, and the work an attached release shipped (task:<id>, bug_report:<id>).
writtenThe newest completed run wrote a highlight for it; false means the fallback.
CustomerEmailDraftDTO
subjectbodyMarkdown.
editedA person's edit; false means the draft composed from the highlights.
The edit's updatedAt (null for a composed draft): send back as expectedDraftUpdatedAt / expectedUpdatedAt.
newerDraftAvailableA run completed after the edit: Use the new draft.
The email's document, a person's edit: its images are filed under this id. Null for a composed draft.
privateImageCountImages in the text that are not public: they are left out of the email when it is sent.
CustomerEmailSendStateDTO
statesentCountCounted sends: send back as expectedSentCount.
The newest attempt.
The newest counted send.
CustomerEmailSkippedDTO
reasonitemKeyitemTitleCustomerEmailDocumentDTO
idThe document's id: the customer-email room's subject, and its comments' entityId.
editedA person edited the email; false while it is the composed draft.
CustomerEmailSendResultDTO
sentfailedrefusedCustomerEmailSendLogPageDTO
totalpagelimitCustomerEmailSendLogRowDTO
idrecipientNamerecipientEmailsubjectstatusdeliveryNumbersendAgaincreatedAtCustomerEmailSendDetailDTO
idrecipientNamerecipientEmailsubjectstatusdeliveryNumbersendAgaincreatedAtbodyExactly the text that was sent (markdown).
MarketingDocumentDTO
kindWhich Marketing document: release-notes (one of a Marketing release's public release notes), blog-post (one of its blog posts) or customer-email (one customer's email about it).
idThe document's id: pass it back to propose-marketing-document-content.
releaseIdThe id of the Marketing release the document belongs to (never a technical release).
releaseNameThe Marketing release's title.
The blog post's title, or the customer the email is for; null for the notes.
markdownThe document's current text, exactly as it stands. Copy edits[].find passages from it.
generationInFlightA new draft is being generated: it will replace the whole text, so no suggestion is taken until it lands.
A suggestion already waiting for review on this document: no second one is taken until a person settles it.
The Marketing release's number: MKT-<releaseDisplayId>.
MarketingDocumentProposalDTO
idsubjectKindWhich Marketing document: release-notes (one of a Marketing release's public release notes), blog-post (one of its blog posts) or customer-email (one customer's email about it).
subjectIdreleaseIdThe id of the Marketing release the document belongs to.
statusbaseMarkdownThe document's text the suggestion was made against.
proposedMarkdownThe whole document as the suggestion would leave it.
reviewCommentIdscreatedAtSandboxDTO
idworkspaceIdnamesizeSandbox size: small (2 vCPU, 8 GB), medium (4 vCPU, 16 GB) or large (8 vCPU, 32 GB)
applicationIdsisDefaultdisplayIddescriptioncreatedByIdcreatedAtupdatedAtWorkspaceEnvVarDTO
keysecretvalueisSetApplicationDTO
idworkspaceIdnamellmAssistedPreviewSetupdisplayIddescriptionrequireDependencyInstallreadOnlyactiveScanscreatedByIdcreatedAtupdatedAtApplicationFromTemplateDTO
idworkspaceIdnamellmAssistedPreviewSetupdisplayIddescriptionrequireDependencyInstallreadOnlyactiveScanscreatedByIdcreatedAtupdatedAtTemplateCopyReport
businessRulesCopiedApplicationEnvironmentDTO
environmentIdkeynametierThe environment's tier: production or non_production.
isDefaultThe branch whose merges deploy to this environment: the application's own, else, for Development and Production, the workspace's Git workflow default. Null when a custom environment is unmapped.
sourceWhere the branch comes from: own, workspace_default or unmapped.
The cloud account this environment deploys to, or null when none is attached.
The environment's deploys to its cloud account: the newest run, the one that runs now, the app's address, and whether a deploy can be retried or started. Null while the environment is attached to none and has no deploy recorded.
ApplicationEnvironmentAttachmentDTO
statusWhere the environment's cloud account attachment is: requested (recorded, nothing set up yet), wiring, wired (deploying), rewiring after a branch change, paused while the account is frozen, detaching, or failed (it can be retried or detached).
kindmanaged: an AWS account Productmind created for the workspace; byo: one the company owns and connected.
tierThe account's tier, always the environment's own.
branchThe branch the environment deploys from, which its deploy role trusts. During a branch change it is the new branch, which the application's own branch may not name yet.
The attached cloud account's id, for those who may read cloud accounts; null for anyone else.
The attached AWS account, as a person reads it; null for those who may not read cloud accounts.
Why the last step failed, while the status is failed; null otherwise, and for those who may not read cloud accounts.
While the status is failed, what failed, which a retry goes on with: attach (setting up deploys), rewire (a branch change, or following a renamed repository) or detach. Null otherwise, and for those who may not read cloud accounts.
True when the attachment failed and can be retried (not while its repository reads as having left Productmind's hosting: should it read present again, the failure says Productmind could not reach it, and a retry is offered); null for those who may not read cloud accounts.
True while nothing has been set up for the environment outside Productmind yet: a request no step has taken, or one that failed before its first step. A detach then removes the attachment at once. Null for those who may not read cloud accounts.
While the status is paused, why nothing deploys: organization_suspended, or account_not_active (the account does not take deploys now). Null while paused but the account takes deploys again (the environment is routed again within a few minutes), null for any other status, and for those who may not read cloud accounts.
What became of the application's repository: renamed (the deployment is following the new name), departed (it reads as having left Productmind's hosting, so only a detach ends the attachment unless it reads present again), or null. Null for those who may not read cloud accounts.
While the environment's branch change is carried out (rewiring, or failed part way), the branch it deployed from before; null otherwise.
True when a detach of the company's own AWS account failed at its last step, removing the deploy role (Productmind can no longer reach the account, could not remove the role there, or failed for another reason), and can be closed, leaving the deploy role for the company to delete. When the repository had left Productmind's hosting, its deploy settings were not removed from it either. Null for those who may not read cloud accounts.
While the detach can be closed, the deploy role it leaves in the company's own AWS account; null otherwise, and for those who may not read cloud accounts.
While the detach can be closed: true when the application's repository had left Productmind's hosting, so its deploy settings stay on it as well as the deploy role in the account. Null otherwise, and for those who may not read cloud accounts.
The environment's domain: the name its app., api., ws. and docs. hosts live under, whether public DNS delegates it yet, and whose it is; for those who may read cloud accounts, a domain the company owns also carries its DNS zone's name servers and the checks. Null while it has none.
A domain change being prepared: the domain the environment moves to once it points at its DNS zone, as domain names one. Its hosts move at the first deploy after its deploy settings name it (domainPairAt). Null with no change pending.
When the environment's domain last changed (ISO 8601). Its hosts move to it at the first deploy after its deploy settings name it (domainPairAt). Null while its domain never changed.
When the repository's deploy settings came to name the environment's changed domain (ISO 8601): a deploy started after it moves the environment's hosts there. Null while its domain never changed, and after a change while the settings are still being written.
The environment's destroy: where one stands, and whether one is offered. Null while destroys are being set up after an update.
ApplicationEnvironmentDestroyDTO
Where a destroy of the environment's stack stands: scheduled (a production environment's destroy waiting its day, still deploying), requested (under way: deploys stopped), restoring (a cancel putting it back in deploys), failed, or destroyed (its stack removed; Deploy again builds it anew). Null with none.
When a scheduled destroy starts (ISO 8601), unless cancelled first. Null otherwise.
When the destroy was asked for, or last tried again (ISO 8601). Null with none.
When the stack was last removed (ISO 8601). Null while it never was.
detachAfterWhether the environment is detached from its cloud account once its stack is removed.
Who asked for the destroy, or last tried it again. Null for a reader who may not read cloud accounts, and with none.
Why a failed destroy failed, as a code. Null unless failed, and for a reader who may not read cloud accounts.
What a failed destroy's failure means and what to do, in words. Null unless failed, and for a reader who may not read cloud accounts.
Whether the environment can be destroyed now: deploying from its cloud account, with no destroy standing and no domain change under way. Null for a reader who may not read cloud accounts.
Whether the environment's stack may run in this cloud account: a deploy of it there did more than not deploy, and no destroy removed the stack since. Null for a reader who may not read cloud accounts.
Whether cancelling the destroy now also deploys the environment's branch, to put back what a failed destroy may have changed in the stack: true for such a failure, false where a cancel only puts the environment back in its deploys (scheduled, under way, or a failure that changed nothing). Null when no destroy can be cancelled now, and for a reader who may not read cloud accounts.
Whether a failed destroy can be tried again now: deploying from its cloud account, no domain change under way, never a cancel GitHub refused, and none of the refusals a press always meets (a departed repository, a switched domain not deployed since, the account taking no deploys, a destroy run of the environment still queued or running, the organization's deletion scheduled). False where none applies. Null for a reader who may not read cloud accounts.
Whether a destroyed environment can be built again now, as far as the environment and its account go: deploying from its cloud account, no domain change under way, and none of the refusals a press always meets. The attach rollout and the plan's Cloud accounts capability are the route's own checks. False where none applies. Null for a reader who may not read cloud accounts.
EnvironmentDeploymentBlockDTO
The environment's newest run, deploy or destroy; null before the first.
The newest run that deployed, which is what runs now; null before the first, and after a run that destroyed the stack.
The environment's app, once a run deployed it and its domain is ready: its address follows the environment's own host settings.
retryableTrue when the newest run is a failed or cancelled deploy that can be run again.
deployableTrue when the environment's branch can be deployed now.
readOnlyTrue once the repository left Productmind's hosting: the records stop there.
EnvironmentDeploymentDTO
idkinddeploy: a run of the repository's deploy workflow; destroy: a run of its destroy workflow.
statusWhere the run is: queued (waiting for its turn or a runner), in_progress, or completed.
GitHub's conclusion of the run, once completed.
Null while the run is not completed, or its result is still being read.
settledTrue once Productmind has read what the completed run did. Until then a completed run's outcome is null, or provisional: a failure whose jobs are still to be read.
runIdGitHub's id of the run.
runNumberThe run's number in the repository's runs of its workflow.
runAttemptThe attempt: a retry is a new attempt of the same run.
latestAttemptFalse when a later attempt of the same run is recorded.
eventWhat started the run: push (a merge into the branch) or workflow_dispatch (Deploy now).
headBranchheadShaThe commit the run deployed.
The job that failed, by its name in the workflow.
The step that failed, by its name in the workflow.
The run on GitHub. Only Productmind support can open it while Productmind hosts the repository.
The person whose Retry or Deploy now started this attempt, when Productmind started it.
runCreatedAtWhen GitHub created the run (ISO 8601).
ApplicationReadinessReportDTO
idworkspaceIdapplicationIdstatushasCompletedRuncreatedAtupdatedAtApplicationDiscoveryRunDTO
idworkspaceIdapplicationIdobjectivestatushasCompletedRuncreatedAtupdatedAtApplicationDocumentationRunDTO
idworkspaceIdapplicationIdstatushasCompletedRuncreatedAtupdatedAtDetectableRepositoryDTO
workforceConnectorIdexternalRepoIdrepoNamerepoUrlproviderbaseBranchownerprivateApplicationDetectionRunDTO
idworkspaceIdstatuscreatedAtupdatedAtScanEstimateDTO
repositoriesThe repository count the estimate is for.
The estimate in AI credits: median and 90th percentile.
The estimate in euro cents at the plan's price per 1M AI credits, rounded up to the cent. For a user-hosted scan (userHosted), the plan's session fee of every session the scan starts instead, since it draws no AI credits. Null when the plan has no such price.
walletBalanceCentsThe organization's pay-as-you-go balance in euro cents. May be negative.
coversMedianWhether a scan of this size may start now: the balance covers the median, or the scan is not charged from the balance.
userHostedThe scan would run on the caller's own Productmind agent, which uses no AI credits (a session fee applies per session).
Why a scan of this size would be refused right now, or null.
ScanStepRefusal
The billing error code the start would get. Null for a refusal no billing action lifts: your organization's own model provider key cannot run the scan.
messageThe sentence that refusal carries.
The amounts behind a balance refusal (what the step needs, the balance, the top-up that covers it), or null.
ResumeWorkspaceScanResponseDTO
The paused scan steps that were started again.
WorkspaceScanRunDTO
idworkspaceIdobjectivestatuscreatedAtupdatedAtLinkedFeatureDTO
idtitlestatuspriorityApplicationRepoPathsDTO
truncatedtotalApplicationApiSpecsDTO
truncatedotherSpecsAPI description files no service took, best first
ApplicationApiSpecService
serviceNameThe port the service serves on in a live preview
pinnedMissingThe pinned file is not in the repository at the base branch
ApplicationApiSpec
pathThe file's path in the repository
sourcepinned by a person, or detected by the rule set
ApplicationApiTargetsDTO
ApplicationApiPreviewTarget
sessionIdnameThe session's title
portbaseUrlThe preview's address for that port, without the API's own base path
ApplicationApiEnvironmentTarget
environmentIdnametierproduction or non_production
liveUrlThe environment's live address
domainRequests may go to this host or any host under it
EnvironmentDTO
idkeynametierThe environment's tier: production or non_production.
isDefaultDevelopment or Production: never deleted, and its tier never changes.
createdAtupdatedAtCloudAccountDTO
idkindmanaged: an AWS account Productmind created for the workspace; byo: one the customer owns and connected.
tierproduction or non_production: only environments of the same tier can use the account.
statusWhere the account is in its life. A managed account goes requested, creating, created, baselining, configuring, then ready; a connected one starts unverified. failed can be retried; retention, queued and closed follow a deletion; closed, handed_over and disconnected are final.
liveFalse once the account is closed, handed over or disconnected; such an account is history.
The 12-digit AWS account id, once known.
The AWS region the account deploys to, once chosen.
Why the last step failed, while the status is failed.
verifyingTrue while Productmind is checking a connected account; it then becomes ready or failed.
cancellableTrue for a request for an account managed by Productmind that has not started: deleting it cancels the request.
The version of the template the account's stack was created from, once checked.
The default version of the account's workload boundary policy, recorded when it was last checked.
When Productmind last checked the account and found it ready.
The version of the terms accepted when the account was added.
A managed account's own network, once its baseline has made it.
When a deletion's retention ends; until then the deletion can be cancelled.
When the account left Productmind, after a hand-over.
When a connection to an account the company owns was disconnected, and Productmind deleted its ExternalId from its records.
Since when a managed account has been frozen over its spend cap: what runs in it keeps serving, and nothing new can be deployed, created or started until this month's spend is under the cap. Null when it is not, and for an account the company connected itself.
createdAtupdatedAtThe environments that deploy to the account (live attachments), on a read of one account or the list; absent from a write's answer.
productmindAddressOn the list alone: whether an environment attached to the account can take Productmind's address, a name under Productmind's own domain, now. False for an account the company connected itself, which needs a domain the company owns, and while Productmind's address isn't available to the organization; a domain the company owns works on either kind. Attaching decides: this says what it would answer. Absent from a read of one account and from a write's answer.
CloudAccountAttachmentRefDTO
idThe attachment's id.
applicationIdThe application whose environment deploys to the account. It may have been deleted since.
environmentIdThe environment's id.
environmentKeyThe environment's key, which never changes.
statusWhere the environment's cloud account attachment is: requested (recorded, nothing set up yet), wiring, wired (deploying), rewiring after a branch change, paused while the account is frozen, detaching, or failed (it can be retried or detached).
awaitingDomainTrue while the environment is being set up and waits for a domain the company owns to point at its DNS zone: the company's DNS change is next.
CloudAccountCostEstimatesDTO
cloudAccountIdmonth^\d{4}-(0[1-9]|1[0-2… · requiredThe month answered, YYYY-MM in UTC.
kindestimate: the current month, or a closed month AWS has not closed yet; statement: the month's amount, fixed once AWS closed the month.
stateFor an estimate: estimated, no_costs, not_read_yet, or not_estimated (an account the company connected itself, which AWS bills directly, or a managed account that has no AWS account yet). For a statement: pending (not charged yet), charged, invoiced, zero (nothing to charge) or held (not charged for now; note says why).
Why a held month is not charged, in words: AWS's figures for the month need a review (no total is shown until they are settled), or the account was handed over to the company. Null for every other month.
When the figures were read from AWS (an estimate), or when the statement was made.
The days the account has been the workspace's, from and to (null while it still is).
currencyThe month's amount in euros, rounded up to the cent; null when there is nothing to convert yet, or for a month held for review (note says why).
The last day the figures cover.
What the total is computed from. A statement's are fixed when it is made; an estimate's are today's plan and the latest rate.
How a statement was charged: from the balance or on an invoice. Null until it is charged.
The total per day, in order; the days add up to the total.
The total per AWS service, largest first; they add up to the total. A service with nothing to charge is left out.
The total per application environment, largest first, with the unallocated part; they add up to the total.
Every month the account has figures for, newest first.
The account's spend cap for the open month: its cap and where its ceiling comes from, this month's figure, the warning's threshold, and whether it is frozen over the cap. Null for any other month, an account the company connected itself, a managed account with no AWS account yet, one queued for closure or closed, or while caps are being set up.
CloudAccountSpendCapDTO
monthThe month the figure is of, YYYY-MM in UTC: the open month.
cap^\d+\.\d{2}$ · requiredThe effective cap in euros: the account's own cap, never above its ceiling.
ceiling^\d+\.\d{2}$ · requiredThe most an Owner or Admin may set, in euros.
ceilingSourceWhere the ceiling comes from: plan (the plan's value for the account's tier), organization (a value Productmind set for the organization), account (a value Productmind set for this account), or default (the platform's default for the tier).
The cap an Owner or Admin set, in euros; null when the account follows its ceiling.
This month's AWS spend in euros as charged, excluding VAT, as the Costs section prices it; null when there is no rate to convert with yet.
Where the figure comes from: estimate, the daily cost estimates; null before the month's first read.
When the figure was read from AWS; null before the month's first read.
The figure at which the Owner and Admins are warned: 80% of the cap; null for a cap of 0, which no warning precedes.
warnedWhether this month's warning was given.
heldWhether the account is frozen over its cap now: nothing new deploys, is created or started until the month's spend is under the cap.
What Productmind limited in the account to stop its spend, while it stays limited.
AuditLogListResponse
Pass back as cursor for the next page; null on the last page.
The Athena execution the page came from (result reuse); null for the local store.
windowFromThe event-time window the read was bounded to, ISO-8601 — 90 days ending at to unless from was given.
windowToAuditEventListItemDTO
id^([0-9a-fA-F]{8}-[0-… · requiredThe event id (event_id).
eventTimeWhen the change committed, ISO-8601 UTC.
operationIdShared by an attempted/completed pair and by every subject of one bulk operation.
sequence0-based order inside operationId.
phasecompleted for every ordinary change; attempted / failed bracket a destructive operation — an attempted with no completed means the process died mid-operation.
actionentityTypeThe audited entity type (registry key), e.g. task.
Writer-formatted, e.g. TASK-12. Null only where the row never had one — a create carries it, read back from the database in the capturing flush.
The entity's title/name at the time of the change.
scopeNull for org-level governance rows (roles, memberships, SSO), which every workspace view of the org shows.
actorKindbrowser = a person; agent = a delegated coding-session token; service = a machine caller; impersonated = a service acting on behalf of a user; support = a Productmind Architect through a support grant (actorTokenId is the grant id).
onBehalfOfThe automation run whose agent made the change, when one did.
sourceThe matched route PATTERN (/workspaces/:workspaceId/tasks/:id), never the concrete path.
changedFieldsThe field names present in changes, after redaction.
Per-field before/after, keyed by field name — inline, so the detail view needs no second read.
changesTruncatedTrue when the record was too large and changes was dropped to {}.
{ entityType: count } of dependants removed by a destructive operation.
producerapi@<version>.
actorRedactedTrue when this organization's view withheld the actor's email and avatar because they are not one of its own people - a Productmind Architect. The page badges exactly these actors.
AuditChangeDTO
kindfromtofrom_lengthto_lengthtruncatedPullRequestCheckFailureDTO
externalIdnamelogTruncatedArchitectureGraphDTO
ArchitectureNodeDTO
idnameisApplicationfeatureCountArchitectureEdgeDTO
idsourceTypesourceIdtargetTypetargetIdServiceFeatureDTO
idtitlekindstatussourceFeatureMapScanModeDTO
proposesTrue when the run writes nothing and proposes its changes for review.
Why it proposes; null when it writes directly.
scannedBeforeTrue when a catalogue scan has run here before: the run is a re-scan, not a first scan.
FeatureMapProposalSummaryDTO
idstatusarmhiddenCountsuppressedCountHow the scan connected the existing features to what it found. Null on a proposal made before this was recorded.
createdAtFeatureMapProposalDTO
idstatusarmhiddenCountsuppressedCountHow the scan connected the existing features to what it found. Null on a proposal made before this was recorded.
Existing features the scan matched to none of its code, limited to the ones this reader may list.
unmatchedHiddenCountUnmatched features left out of the list: private to this reader, or past the list's limit.
createdAtFeatureMapDeltaDTO
idtypestatusFEAT-n, COMP-n or DOM-n of the feature the change is about.
The feature's title now, which a move or merge row names it by.
The feature's provenance marker (repo_discovered for one a scan made), for its leading glyph.
A component's layer (view, API service, data repository), for its leading glyph; null on a feature or domain.
evidenceeditedByPersonchangedSinceScanThe feature changed after the scan read it: its title, place or status no longer matches the row's before value.
carriedOverThe row this one follows: an addition's parent addition, else the row of the same type on the nearest feature above its own. Null when it stands on its own.
person for a decision made on this row, inherited for one it follows from the row above; null while undecided.
An inherited decision's origin: the row whose decision it follows.
Tags held on the change. They land on its feature when it is applied accepted or amended.
Who the change is assigned to: people and teams.
person, inherited (from a row above a person assigned) or owner (its feature's owners); null when unassigned.
An inherited assignment's origin: the row a person assigned.
assignedToViewerAssigned to the reader, or to a team they belong to.
commentCountComments in the change's thread.
FeatureGraphDTO
FeatureGraphNodeDTO
idtitledepthstatuspriorityisAiGeneratedkindspecificationStateLatestWorkspaceScanRunDTO
idworkspaceIdobjectivestatuscreatedAtupdatedAtWhen a scan of this kind last completed; null when none ever has.
SessionScopePreflight
readyTrue when the session would clone at least one repository and has a sandbox and somewhere to run: the Productmind cloud, or, while cloud sessions are turned off for the workspace, the person's own Productmind agent.
Empty when ready.
The sandbox the session would run in - the caller's choice, else the workspace default.
Every application in the workspace, so the user can be offered the ones not yet in scope.
What the session would actually clone, deduplicated.
The sandboxes to choose from. Present only for a caller who may read them, and only when the choice is open.
SessionScopeBlocker
codemessagePlain-language statement of what is missing and who can fix it — relay it to the user.
SessionScopeApplication
idnameattachedToSandboxSessionScopeRepository
repoNameapplicationIdviareadOnlyTrue when an application on this repository is read-only: the session would read it for context and never push to it or open a pull request on it.
SessionMenuItemDTO
id^([0-9a-fA-F]{8}-[0-… · requirednamecategorystatusSessionRunnerDTO
keycloud, mcp, none, or the Productmind agent's UUID
The build mode behind it; null for sessions not started yet
The Productmind agent's UUID, for an agent
The Productmind agent's name, as its most recent session recorded it
AddSessionItemsResult
messageSessionSuggestedRepository
applicationIdapplicationNamerepoNameSessionTaskDTO
iddisplayIdtitledescriptionstatusSessionFeatureDTO
iddisplayIdtitledescriptionrestrictedSessionBugReportDTO
iddisplayIdtitledescriptionseveritystatusprioritySessionBugReportReporterDTO
namedetailsThe reporter's own words — usually the repro steps
SessionBugReportRelatedEntityDTO
entityTypefeature | application | task | …
entityIddisplayIdtitlerelationshipTypeSessionBugReportConvertedTaskDTO
iddisplayIdtitlestatusDraftPlanFromChatResponse
Decision Table
| Variant | Matching Criteria |
|---|---|
| type = object · requires: promoted, sessionId | |
| type = object · requires: scheduled, sessionId |
promotedsessionIdSessionPushAccess
verdictcheckedAtrunKindSessionPushAccessRepository
repoNameworkingBranchverdictreasonsentencepusherSessionPushAccessRule
sourcenameruleeffectdetailpatternrequiredAccessLevelSessionCostDTO
currencyISO 4217 currency of every amount, lower case
chargedCentsEverything the session was charged so far, in cents: AI credits, sandbox time and the session fee
The session's AI credits
Sandbox time; a preview still running is charged when it stops
The fee for a session built on your own machine, in cents (0 when your plan includes it); null when there is none
billedByInvoiceTrue when some of this cost is billed on your organization's invoice instead of its credit balance
floorTrue when a turn is still running, or ended before reporting its final usage: the totals are a lower bound
Every turn that drew AI credits, oldest first
SessionCostTurn
keyStable id of the turn within the session
labelWhat the turn did: plan, build, documentation, pr_review, live_edit, publish_live, scan, or another step name
firstAt^(?:(?:\d\d[2468][04… · requiredWhen the turn's first usage was recorded
lastAt^(?:(?:\d\d[2468][04… · requiredWhen the turn's latest usage was recorded
creditsAI credits the turn drew
chargedCentsWhat the turn's AI credits were charged, in cents
completeFalse while the turn is still running, or when it ended before reporting its final usage (its figures are a lower bound)
SessionGitIdentityAccount
providerconnectedSessionGitIdentityRepository
repoNameprovideractorSessionTranscriptUploadDTO
uploadIdPass this to import-session-transcript once command has run.
uploadUrlWhere command uploads the transcript. It expires at expiresAt; you do not need to use it yourself.
expiresAtISO 8601 time the upload URL stops working. Request a new upload after it.
maxBytesThe largest upload the import accepts, in bytes.
commandRun this verbatim with your shell tool. It finds your own conversation's transcript on this machine (the one file holding uploadId) and uploads it. The person may be asked to approve it; it needs network access.
SessionTranscriptImportResultDTO
clientThe transcript format that was read.
completeFalse when the import stopped at its time limit: call import-session-transcript again with the same uploadId until it is true.
importedConversation entries added to the session by this call.
skippedEntries the session already had from an earlier share, left as they were.
imagesImages from your conversation (screenshots in tool results, images in prompts) that the session now shows. When above zero, do not upload those screenshots again.
omittedImagesImages that could not be stored (too large, unsupported type, or the session's attachment limit).
conversationIdsYour client's own id for the conversation that was read.
SessionRunErrorList
totalfatallastRunStoppedSessionRunError
idsessionIdcodeseveritystagetitlemessagedetailsHiddenapplicationIdssourceoccurredAtSessionRunErrorBodyAccess
urlexpiresAtISO 8601 time the URL stops working
SessionSubagentList
runningtotalSessionSubagent
idThe sub-agent's id in Productmind
sessionIdproviderexternalIdThe provider's id for it: Claude's task id, Codex's thread id
The sub-agent that started this one; null when the main agent did
The tool call that started it
statusstartedAtresumeCountmessageCountSessionSubagentAvailability
providerstatereportedAtSessionSubagentDetail
idThe sub-agent's id in Productmind
sessionIdproviderexternalIdThe provider's id for it: Claude's task id, Codex's thread id
The sub-agent that started this one; null when the main agent did
The tool call that started it
statusstartedAtresumeCountmessageCountIts final report, markdown
What went wrong, when it failed
SessionSubagentMessage
idsubagentIdsessionIdpositionsourcekindcontentcreatedAtupdatedAtSessionPullRequestDTO
urlsourcebehindByCappedsyncSupportedPullRequestCommitsResultDTO
SessionDocumentationContextDTO
markdowndraftingInstructionsSessionDocumentationContextApplicationDTO
idnameworkingDirectoryhasTechnicalDocsdocIndexMarkdownSessionDocumentationContextFeatureDTO
idtitlehasSpecificationSessionConversationDTO
sessionIdcontenttruncatedchatIdSessionPreviewShareDTO
availablecanManageenabledexpiryexpiredcommentsEnabledaiRepliesEnabledCodingErrorsSummary
daysCodingErrorsList
SettingDTO
iddomainnamespacenamevalueTagDTO
identityentityIdnamesourceWho applied the tag: user (a person or API caller) or generated (stamped by the platform)
statusactive — the entity carries the tag. suggested — proposed by the tagger, inert until a person accepts it (only returned with includeSuggested=true). rejected rows are never returned.
The tagger's one-line justification for a suggested tag
createdAtupdatedAtTagGenerateResponseDTO
queuedHow many tag.generate messages were published
entityIdsTagSuggestionsAcceptResponseDTO
acceptedTagSuggestionNewNameDTO
namecountHow many entities of the type the tagger proposed it on
TagSuggestionByNameResponseDTO
affectedHow many suggested rows were accepted / rejected
GitBranchPresenceDTO
providerexternalRepoIdThe GitLab server the repository is on, as its key names it; null for gitlab.com, GitHub and Bitbucket
Whether the repository has the branch; null when its branches could not be read
GitConnectionAccess
The connection's provider, read from its Applications; null when none use it
supportedWhether this provider's connection can be checked yet (GitLab)
The account every action on this connection runs as; null when nothing was checked
Why the check could not run, e.g. gitlab_connection_refused: the connection must be reconnected
GitConnectionWebhooks
The connection's provider, read from its applications; null when none use it
supportedWhether Productmind manages this provider's webhooks (GitLab, Bitbucket)
The URL every webhook of this connection delivers to; null when the provider was not asked
One per project, in the order of its first application's name
Why the provider could not be asked about every project: gitlab_connection_refused (reconnect the connection), timeout, or another failure
GitConnectionWebhookProject
externalRepoIdThe GitLab project's id, or the Bitbucket repository's UUID
repoNameThe project's path (group/name, or workspace/slug on Bitbucket), as its applications record it
The project's URL, as its applications record it
The applications that use the project: a monorepo's are several
statuspresent: a webhook delivers merge requests (pull requests on Bitbucket) to this connection. stale: it delivers to a connection of this workspace that was replaced. missing: none delivers to this workspace. added / repointed: Productmind just added one, or moved the stale one onto this connection. no_permission: the connection's account cannot manage the project's webhooks (on GitLab it takes Maintainer; on Bitbucket, repository admin and a connection allowed to manage webhooks), so a person adds it by hand. not_found: the provider does not show the account the project. failed: the provider refused for another reason, in detail. unchecked: the provider could not be asked; error says why.
The provider's id of the webhook, where there is one
disabledGitLab turned the webhook off after failed deliveries. A successful test request from the webhook's page in GitLab turns it back on.
What the provider said, where it refused or failed
GitConnectorRebindResult
applicationsApplication repository blobs repointed
sessionRepositoriesSession git-repository rows repointed
releaseApplicationsRelease application snapshots repointed
notificationChannelsChannel subscriptions repointed
unresolvedRepositories no connected provider can currently serve. Left pointing at their existing connector on purpose — a reconnect that granted access to fewer repositories must surface, not be silently rebound onto a connector that cannot see them.
Present when the workspace has a GitLab connection with projects in use: after the links are repaired, each project's merge request webhook that still delivers to a replaced connection is moved onto the live one
AuditEventDetailDTO
id^([0-9a-fA-F]{8}-[0-… · requiredThe event id (event_id).
eventTimeWhen the change committed, ISO-8601 UTC.
operationIdShared by an attempted/completed pair and by every subject of one bulk operation.
sequence0-based order inside operationId.
phasecompleted for every ordinary change; attempted / failed bracket a destructive operation — an attempted with no completed means the process died mid-operation.
actionentityTypeThe audited entity type (registry key), e.g. task.
Writer-formatted, e.g. TASK-12. Null only where the row never had one — a create carries it, read back from the database in the capturing flush.
The entity's title/name at the time of the change.
scopeNull for org-level governance rows (roles, memberships, SSO), which every workspace view of the org shows.
actorKindbrowser = a person; agent = a delegated coding-session token; service = a machine caller; impersonated = a service acting on behalf of a user; support = a Productmind Architect through a support grant (actorTokenId is the grant id).
onBehalfOfThe automation run whose agent made the change, when one did.
sourceThe matched route PATTERN (/workspaces/:workspaceId/tasks/:id), never the concrete path.
changedFieldsThe field names present in changes, after redaction.
Per-field before/after, keyed by field name — inline, so the detail view needs no second read.
changesTruncatedTrue when the record was too large and changes was dropped to {}.
{ entityType: count } of dependants removed by a destructive operation.
producerapi@<version>.
actorRedactedTrue when this organization's view withheld the actor's email and avatar because they are not one of its own people - a Productmind Architect. The page badges exactly these actors.
The other phases of the same operation (attempted ↔ completed / failed), and the sibling subjects of a bulk operation.
AuditExportStatusDTO
executionIdformatstatusdownloadUrlWhere to fetch the file: a presigned URL valid ~1 h (re-signed on every poll) — or, on the local store, the API's own download route.
downloadUrlsEvery part of a multi-part JSON export; downloadUrl is the first.
expiresAtbytesScannederrorAuditSettingsDTO
Days kept; null = indefinite.
isDefaultTrue when the workspace never saved a value — the registry default applies and follows future default changes; a saved value is returned verbatim forever.
PlanPricingFactsDTO
Always empty: per-model prices are not published.
PublicBlogPostListDTO
revisionThe revision of exactly these posts: what the blog's revision.json publishes once built from them.
Every post on the blog, newest first publish first.
linksExpireAtWhen the short-lived links in this answer stop working.
PublicBlogPostSummaryDTO
slugtitlesummarypublishedAtWhen the post was first put on the blog.
updatedAtWhen the version the blog serves last changed.
PublicBlogHeroDTO
kindurlWhere to fetch it: its address on the assets host when published there, else a short-lived link.
contentTypePublicBlogPostDTO
slugtitlesummarypublishedAtWhen the post was first put on the blog.
updatedAtWhen the version the blog serves last changed.
contentThe post's markdown, as approved, with each of its images given a short-lived link.
linksExpireAtWhen the short-lived links in this answer stop working.
PublicBlogRevisionDTO
revisionA hash of the posts on the blog: each one's slug and the version it serves.
postsHow many posts are on the blog.
__schema0
iddisplayIdworkspaceIdtitlestatustypeprioritysortOrderchildCountdescriptionassigneeIdparentIdstartAtdueAtexternalSourceexternalIdexternalSourceUrltagssuggestedTagCountcreatedAtupdatedAtAgentDefinitionModelOutput
tierModel capability class — quick, standard, top or ultra. Ignored when id is set.
idExplicit model id for the active LLM provider; a Bedrock-style id is normalized on the Anthropic API. Beats tier.
temperaturemaxTokensthinkingEnabledthinkingEffortpromptCachingAgentDefinitionMcpServerOutput
Decision Table
| Variant | Matching Criteria |
|---|---|
| type = object · kind="productmind" · requires: id | |
| type = object · kind="integration" · requires: id, instanceId | |
| type = object · kind="http" · requires: id, transport, url |
idIdentifier for this server within the definition. Never used as a shared cache key.
kindallowedToolsTool-name globs to load from this server. Required for the asynchronous path — an absent list means 'everything this server exposes', which cannot be validated at publish time.
rejectedToolsTool-name globs to exclude, applied after allowedTools and on top of the global rejections.
ProductmindMcpServerOutput
idIdentifier for this server within the definition. Never used as a shared cache key.
kindallowedToolsTool-name globs to load from this server. Required for the asynchronous path — an absent list means 'everything this server exposes', which cannot be validated at publish time.
rejectedToolsTool-name globs to exclude, applied after allowedTools and on top of the global rejections.
IntegrationMcpServerOutput
idIdentifier for this server within the definition. Never used as a shared cache key.
kindinstanceIdWorkforce instance id of an MCP integration already attached to this workspace.
allowedToolsTool-name globs to load from this server. Required for the asynchronous path — an absent list means 'everything this server exposes', which cannot be validated at publish time.
rejectedToolsTool-name globs to exclude, applied after allowedTools and on top of the global rejections.
HttpMcpServerOutput
idIdentifier for this server within the definition. Never used as a shared cache key.
kindtransporturlThe MCP server endpoint. Our infrastructure connects outbound to it, unauthenticated.
allowedToolsTool-name globs to load from this server. Required for the asynchronous path — an absent list means 'everything this server exposes', which cannot be validated at publish time.
rejectedToolsTool-name globs to exclude, applied after allowedTools and on top of the global rejections.
AgentDefinitionSettingsOutput
maxTurnsCap on agent-loop turns. The bound on how long a tool-using run can iterate.
maxTotalTokensCap on tokens processed per run, including cache reads and writes; not cost-weighted.
AgentDefinitionKbFiltersOutput
includeTagsexcludeTagsAutomationTriggerOutput
Decision Table
| Variant | Matching Criteria |
|---|---|
| type = object · requires: type, eventType | |
| type = object · requires: type, frequency, timezone | |
| type = object · requires: type |
typeeventTypeA workspace event this automation reacts to, e.g. bug_report.created.
AutomationConditionsOutput
combinatorAutomationConditionNodeOutput
Decision Table
| Variant | Matching Criteria |
|---|---|
| type = object · requires: field, operator | |
| type = object · requires: combinator, children |
fieldA field declared for this trigger's event by automationConditionFields.
operatorAutomationConditionOutput
fieldA field declared for this trigger's event by automationConditionFields.
operatorAutomationConditionGroupOutput
combinatorAutomationActionOutput
Decision Table
| Variant | Matching Criteria |
|---|---|
| type = object · requires: type, agentSlug, input |
typeagentSlug^[a-z0-9]+(-[a-z0-9]… · requiredThe agent definition to run. Must resolve, be enabled and be async-eligible.
inputThe per-run instruction, as a template over {{entity.id}}, {{entity.key}}, {{entity.title}}, {{entity.status}}, {{event}}, {{actor.name}} and {{changes.<field>.to}} / .from. Unknown placeholders render empty and are recorded as warnings on the run.
AutomationLimitsOutput
maxRunsPerHourRuns per hour for this automation, clamped to [1, 200]. Defaults to 50. Counted with an indexed query, and it FAILS CLOSED.
RunTelemetryOutput
modelCallCounttoolCallCounttoolErrorCountstepsTruncatedstepsDroppedTelemetryStepOutput
kindnamestartedAtMsstatuspreviewOmittedreasoningTruncatedinputTruncatedoutputTruncatedCommentAnchorOutput
Decision Table
| Variant | Matching Criteria |
|---|---|
| type = object · requires: type, startNodeKey, startOffset +3 more | |
| type = object · requires: type, pullRequestUrl, pullRequestNumber +4 more | |
| type = object · requires: type, port, pagePath +4 more |
typestartNodeKeystartOffsetendNodeKeyendOffsetsnippetprefixsuffixtextStarttextEndFeatureBusinessRuleCitationDTOOutput
labelrepoNameshapathBugReportProposedRelationshipOutput
targetTypetargetIdconfidenceSessionInteractionOptionOutput
valuelabeldescriptionrecommendedSet true on the option you recommend (put it first). The app labels it Recommended and picks it in advance, so the user can accept it with one click. Omit it when you have no recommendation; on a single-choice question flag at most one.
PreviewPortOutput
portnameprotocolisMainInternalMeetingActionPointOutput
descriptiondoneassigneeUserIdassigneeEmailBlogPostGuidanceOutput
tonelengthangleA free-text angle for the post, e.g. who it is for or what to lead with.
SidecarEnvOutput
key^[A-Za-z_][A-Za-z0-9… · requiredcommandApplicationRiskRuleOutput
idUnique within the profile
nameShown as the reason a file is flagged
tierhigh and medium flag a file; a file matching any low rule is never flagged
pathPatternsGlobs over the repository-relative path (*, **, ?, {a,b}), case-insensitive. A pattern with no / matches the file name in any folder
nameWordsWhole words of a folder or file name, case-insensitive. Names split on punctuation and camelCase, and two adjacent pieces also match joined
lineTextsText an added or removed line contains as whole words, case-insensitive, with any run of spaces between words
Matches when additions plus deletions reach this; null for no threshold
ApplicationRepoOutput
workforceConnectorIdexternalRepoIdrepoNamerepoUrlproviderbaseBranchApplicationServiceOutput
nameThe service's name, e.g. web or api
typeWhat kind of unit the service is
portThe port it serves on in a live preview; leave it out for a service with nothing to preview
protocolHow the port is reached: http (the default) or https
isMainWhether this port is the one a live preview opens first
The service's OpenAPI or Swagger file, a path in the repository such as services/api/openapi.json. Leave it out to keep the stored one; null clears it, so the file is found automatically again