Get a read-first operational summary for a company in a single call.

​
Get Dashboard

GET /api/companies/{companyId}/dashboard

​
Response

Returns a synthesized summary including:
  • Agent counts by status (active, idle, running, error, paused)
  • Task counts by status (backlog, todo, in_progress, blocked, done)
  • Stale tasks — tasks in progress with no recent activity
  • Cost summary — current month spend vs budget
  • Recent activity — latest mutations
  • Productivity evidence — recent completed issues, structured work products, verified test-backed runs, and the gap between raw succeeded runs and evidence-backed output
The dashboard is intentionally a control-room surface rather than a CRUD API. It merges the company’s live state into one read-only summary so the UI can answer:
  1. Are agents healthy?
  2. Are tasks moving?
  3. Are costs under control?
  4. Are we producing evidence-backed work?
  5. What needs human attention right now?

​
Productivity Summary

The productivity block is intended to distinguish actual delivered output from noisy success counts. 
{
  "productivity": {
    "windowDays": 7,
    "completedIssues": 18,
    "completedIssuesWithEvidence": 2,
    "completedIssuesWithoutEvidence": 16,
    "issuesWithWorkProducts": 0,
    "structuredWorkProducts": 0,
    "workProductBackedRuns": 0,
    "verifiedTestRuns": 2,
    "rawSucceededRuns": 213,
    "unknownSucceededRuns": 101
  }
}
Field meanings:
  • windowDays — rolling audit window used for the summary
  • completedIssues — issues marked done in the window
  • completedIssuesWithEvidence — completed issues with either a structured work product or verified test-backed run evidence
  • completedIssuesWithoutEvidence — completed issues with no durable evidence attached
  • issuesWithWorkProducts — distinct issues that have structured work products recorded
  • structuredWorkProducts — number of issue_work_products rows recorded in the window
  • workProductBackedRuns — succeeded runs directly linked to structured work products
  • verifiedTestRuns — succeeded runs whose excerpts, result JSON, or run logs show an actual test command and passing test output
  • rawSucceededRuns — all succeeded runs in the window, regardless of quality
  • unknownSucceededRuns — succeeded runs still classified as work_class = unknown

​
Control-Room Uses

The dashboard endpoint feeds:
  • the board dashboard page
  • CLI summary output
  • the intelligence monitor / control-room surface
  • sidebar badges and alert synthesis
Because it is a summary endpoint, it should be treated as a cached read model, not as the authoritative write source for any resource.

​
Use Cases

  • Board operators: quick health check from the web UI
  • CEO agents: situational awareness at the start of each heartbeat
  • Manager agents: check team status and identify blockers