arazzo: 1.0.1
info:
  title: Sentry Triage and Resolve an Issue
  summary: Find the most relevant unresolved issue, inspect its events, then resolve or reprioritize it.
  description: >-
    A core Sentry incident-response pattern. The workflow lists an
    organization's unresolved issues using a Sentry query, branches on whether
    any issues were returned, retrieves the top issue's detail, pulls its most
    recent error events for context, and then updates the issue to a resolved
    status with a priority. Every step spells out its request inline so the flow
    can be read and executed without opening the underlying OpenAPI description.
  version: 1.0.0
sourceDescriptions:
- name: sentryApi
  url: ../openapi/sentry-api-openapi.yml
  type: openapi
workflows:
- workflowId: triage-issue
  summary: Triage the top unresolved issue in an organization and resolve it.
  description: >-
    Lists unresolved issues for the organization sorted by frequency, selects
    the first match, retrieves its full detail and recent events, then marks the
    issue resolved with an explicit priority. Branches to end when no
    unresolved issues exist.
  inputs:
    type: object
    required:
    - organizationSlug
    properties:
      organizationSlug:
        type: string
        description: The organization slug that owns the issues (e.g. my-organization).
      query:
        type: string
        description: Sentry query string used to filter issues.
        default: "is:unresolved"
      statsPeriod:
        type: string
        description: Time range window for the issue search (e.g. 14d, 24h).
        default: "14d"
      priority:
        type: string
        description: Priority to set when resolving the issue.
        default: "high"
  steps:
  - stepId: findTopIssue
    description: >-
      List the organization's issues matching the supplied query, sorted by
      frequency, returning at most one result to triage first.
    operationId: listOrganizationIssues
    parameters:
    - name: organization_slug
      in: path
      value: $inputs.organizationSlug
    - name: query
      in: query
      value: $inputs.query
    - name: statsPeriod
      in: query
      value: $inputs.statsPeriod
    - name: sort
      in: query
      value: freq
    - name: limit
      in: query
      value: 1
    successCriteria:
    - condition: $statusCode == 200
    outputs:
      topIssueId: $response.body#/0/id
      topIssueShortId: $response.body#/0/shortId
    onSuccess:
    - name: issueFound
      type: goto
      stepId: getIssueDetail
      criteria:
      - context: $response.body
        condition: $.length > 0
        type: jsonpath
    - name: noIssues
      type: end
      criteria:
      - context: $response.body
        condition: $.length == 0
        type: jsonpath
  - stepId: getIssueDetail
    description: >-
      Retrieve full stats and metadata for the selected issue so the triage
      decision is made against current data.
    operationId: retrieveIssue
    parameters:
    - name: organization_slug
      in: path
      value: $inputs.organizationSlug
    - name: issue_id
      in: path
      value: $steps.findTopIssue.outputs.topIssueId
    successCriteria:
    - condition: $statusCode == 200
    outputs:
      issueTitle: $response.body#/title
      issueLevel: $response.body#/level
      issueCount: $response.body#/count
  - stepId: listRecentEvents
    description: >-
      Pull the most recent error events bound to the issue, including full
      stack-trace data, to confirm the cause before resolving.
    operationId: listIssueEvents
    parameters:
    - name: organization_slug
      in: path
      value: $inputs.organizationSlug
    - name: issue_id
      in: path
      value: $steps.findTopIssue.outputs.topIssueId
    - name: full
      in: query
      value: true
    - name: limit
      in: query
      value: 5
    successCriteria:
    - condition: $statusCode == 200
    outputs:
      latestEventId: $response.body#/0/eventId
  - stepId: resolveIssue
    description: >-
      Mark the issue resolved and assign an explicit priority so it leaves the
      unresolved queue.
    operationId: updateIssue
    parameters:
    - name: organization_slug
      in: path
      value: $inputs.organizationSlug
    - name: issue_id
      in: path
      value: $steps.findTopIssue.outputs.topIssueId
    requestBody:
      contentType: application/json
      payload:
        status: resolved
        priority: $inputs.priority
        hasSeen: true
    successCriteria:
    - condition: $statusCode == 200
    outputs:
      resolvedStatus: $response.body#/status
  outputs:
    issueId: $steps.findTopIssue.outputs.topIssueId
    issueShortId: $steps.findTopIssue.outputs.topIssueShortId
    issueTitle: $steps.getIssueDetail.outputs.issueTitle
    latestEventId: $steps.listRecentEvents.outputs.latestEventId
    resolvedStatus: $steps.resolveIssue.outputs.resolvedStatus