> ## Documentation Index > Fetch the complete documentation index at: https://docs.signalmd.dev/llms.txt > Use this file to discover all available pages before exploring further. # Get a run's result document > The run as one document (schemaVersion 1), the same the CLI writes and webhooks post: every cell with its attempts, verdicts and who decided them, and the change from the previous comparable run, or from `compareTo`. ## OpenAPI ````yaml /api-reference/openapi.json get /v1/agents/{agentId}/runs/{runId}/result openapi: 3.1.0 info: title: Signal API version: 1.0.0 description: >- Signal reads a voice agent's production calls and says what to fix, then tests whether the fix worked. This API exposes the same data the web app shows: agents, calls and their traces, signals, issues, versions and their comparisons, tests, and test runs. Authenticate with `Authorization: Bearer sig_...` (an API key from the API keys page). Read keys may call every `read` operation; write keys may also call `write` operations, within their monthly budget. Each operation's scope is in `x-signal-scope`. Errors share one envelope (`Error`), lists share one cursor scheme (`limit`, `cursor`, `hasMore`, `nextCursor`), and every response carries `Request-Id` and `RateLimit-*` headers. servers: - url: https://first-signal-api.fly.dev description: Production - url: http://localhost:43124 description: Local development security: - bearerAuth: [] tags: - name: Account description: Who the token belongs to and what it may do. - name: Agents description: Voice agents connected to the workspace. - name: Calls description: Production calls, their traces and per-call signals. - name: Signals description: What Signal noticed across calls. - name: Issues description: 'Problems worth fixing: where, why, and what to change.' - name: Versions description: Agent versions and how each one did. - name: Comparisons description: Stored comparisons between two versions. - name: Tests description: Tests, test sets, personas and coverage. - name: Runs description: Test runs and the attempts inside them. paths: /v1/agents/{agentId}/runs/{runId}/result: get: tags: - Runs summary: Get a run's result document description: >- The run as one document (schemaVersion 1), the same the CLI writes and webhooks post: every cell with its attempts, verdicts and who decided them, and the change from the previous comparable run, or from `compareTo`. operationId: getRunResult parameters: - schema: type: string format: uuid pattern: >- ^[0-9a-fA-F]{8}-[0-9a-fA-F]{4}-[0-9a-fA-F]{4}-[0-9a-fA-F]{4}-[0-9a-fA-F]{12}$ example: 3f1c2d4e-5a6b-4c7d-8e9f-0a1b2c3d4e5f description: The agent's id. required: true description: The agent's id. name: agentId in: path - schema: type: string format: uuid pattern: >- ^[0-9a-fA-F]{8}-[0-9a-fA-F]{4}-[0-9a-fA-F]{4}-[0-9a-fA-F]{4}-[0-9a-fA-F]{12}$ example: 3f1c2d4e-5a6b-4c7d-8e9f-0a1b2c3d4e5f description: The run's id. required: true description: The run's id. name: runId in: path - schema: type: string format: uuid pattern: >- ^[0-9a-fA-F]{8}-[0-9a-fA-F]{4}-[0-9a-fA-F]{4}-[0-9a-fA-F]{4}-[0-9a-fA-F]{12}$ example: 3f1c2d4e-5a6b-4c7d-8e9f-0a1b2c3d4e5f description: Compare with this run instead of the previous comparable one. required: false description: Compare with this run instead of the previous comparable one. name: compareTo in: query responses: '200': description: The result document. content: application/json: schema: $ref: '#/components/schemas/ResultDocument' '400': description: 'Invalid request: a parameter or the body failed validation.' content: application/json: schema: $ref: '#/components/schemas/Error' '401': description: Missing, invalid or revoked credentials. content: application/json: schema: $ref: '#/components/schemas/Error' '403': description: Not permitted for this principal. content: application/json: schema: $ref: '#/components/schemas/Error' '404': description: >- The agent, or a resource under it, does not exist or is not visible to this key. content: application/json: schema: $ref: '#/components/schemas/Error' '429': description: Rate limited. See Retry-After. content: application/json: schema: $ref: '#/components/schemas/Error' '500': description: Server error. The message is generic; quote the request id. content: application/json: schema: $ref: '#/components/schemas/Error' x-codeSamples: - lang: bash label: curl source: >- curl "https://first-signal-api.fly.dev/v1/agents/6f1d2c3b-4a5e-4f60-8b7a-9c0d1e2f3a4b/runs/d1e2f3a4-b5c6-4d7e-8f9a-0b1c2d3e4f5a/result" \ -H "Authorization: Bearer $SIGNAL_API_KEY" components: schemas: ResultDocument: type: object properties: schemaVersion: type: number enum: - 1 batchId: type: string description: >- The run's id (runId on v1). The document keeps its schemaVersion 1 names because CI and webhooks already read them. agentId: type: string status: type: string description: >- The run's status: queued, awaiting_approval, running, complete, refused or failed. message: type: - string - 'null' kind: type: string description: run or calibration. trigger: type: - string - 'null' targetVersion: type: object properties: platform: type: string enum: - retell version: type: integer configHash: type: - string - 'null' required: - platform - version - configHash comparedTo: type: - object - 'null' properties: batchId: type: string description: The compared run's id. version: type: integer required: - batchId - version epoch: $ref: '#/components/schemas/ScoringEpoch' summary: type: object properties: cells: type: integer gating: type: integer passed: type: integer failed: type: integer flaky: type: integer error: type: integer drift: type: integer notRun: type: integer newlyBroken: type: integer fixed: type: integer required: - cells - gating - passed - failed - flaky - error - drift - notRun - newlyBroken - fixed cost: type: object properties: retellUsd: type: number signalUsd: type: number required: - retellUsd - signalUsd description: >- Retell's share is the customer's spend; Signal's is ours to serve. Estimates. cells: type: array items: $ref: '#/components/schemas/ResultCell' shadow: type: object properties: shadowAgentId: type: string clonedFromVersion: type: integer required: - shadowAgentId - clonedFromVersion description: >- Present when the run played a shadow copy; the version is the one it was cloned from. required: - schemaVersion - batchId - agentId - status - message - kind - trigger - targetVersion - comparedTo - epoch - summary - cost - cells description: >- A run as one JSON document, the same one the CLI writes and webhooks post. Its field names are fixed by schemaVersion 1 and predate v1 naming: batchId is the run's id, and attempts[].runId is an attempt's id. example: schemaVersion: 1 batchId: d1e2f3a4-b5c6-4d7e-8f9a-0b1c2d3e4f5a agentId: 6f1d2c3b-4a5e-4f60-8b7a-9c0d1e2f3a4b status: complete message: null kind: run trigger: ci targetVersion: platform: retell version: 8 configHash: b41f9c0e2a7d comparedTo: batchId: e5f6a7b8-c9d0-4e1f-9a2b-3c4d5e6f7a8b version: 7 epoch: simulatorModel: claude-haiku-4-5 judgeModel: claude-sonnet-4-5 judgePromptHash: 9f2c1a7e scorerVersion: '2026-09-18' summary: cells: 12 gating: 10 passed: 9 failed: 2 flaky: 1 error: 0 drift: 0 notRun: 0 newlyBroken: 0 fixed: 2 cost: retellUsd: 1.84 signalUsd: 0.21 cells: - testId: 3f1c2d4e-5a6b-4c7d-8e9f-0a1b2c3d4e5f testRevision: 2 source: production persona: '@source' tier: text lane: integrations expect: pass guard: false gates: true gateReason: Calibrated. result: passed attempts: - runId: f9a0b1c2-d3e4-4f5a-8b6c-7d8e9f0a1b2c attempt: 1 result: passed k: 1 'n': 1 change: fixed reason: null criteria: - criterionId: c1 text: The agent offers at least one time for a cleaning met: true decidedBy: kind: judge model: claude-sonnet-4-5 answers: [] evidenceTurn: 9 quote: I have Tuesday at 10 or Thursday at 2. noFixture: [] retellGrader: status: null explanation: null findings: [] cost: retellUsd: 0.15 signalUsd: 0.02 Error: type: object properties: error: type: object properties: type: $ref: '#/components/schemas/ErrorType' code: type: string example: agent_not_found description: The specific reason; stable within v1. message: type: string example: Agent not found. param: type: - string - 'null' description: The request field at fault, for a validation error. requestId: type: - string - 'null' example: req_8c1f0e2b4a6d9e3f7a5c1b0d quota: type: object properties: metric: type: string example: test_calls description: analysed_calls, agents, copilot_messages or test_calls. limit: type: integer used: type: integer plan: type: string example: free_org description: The workspace's plan slug. upgradePath: type: string example: /billing description: Where in the web app the plan is changed. required: - metric - limit - used - plan - upgradePath description: >- Present when code is quota_exceeded: the plan limit the request would pass. required: - type - code - message - param - requestId required: - error example: error: type: not_found_error code: agent_not_found message: Agent not found. param: null requestId: req_8c1f0e2b4a6d9e3f7a5c1b0d ScoringEpoch: type: - object - 'null' properties: simulatorModel: type: - string - 'null' description: What played the caller. judgeModel: type: - string - 'null' description: What scored the criteria. judgePromptHash: type: - string - 'null' description: >- Identifies how results were scored. Results with different values were scored differently. scorerVersion: type: - string - 'null' description: The version of Signal's scoring. required: - simulatorModel - judgeModel - judgePromptHash - scorerVersion description: >- What scored a result. Results scored in different epochs are not compared. ResultCell: type: object properties: testId: type: string description: The test. testRevision: type: - integer - 'null' description: The test revision its attempts ran. source: type: string description: production, custom or probe. persona: type: string description: The persona slug, or @source for the test's own caller. tier: type: string lane: type: - string - 'null' expect: type: string enum: - pass - fail guard: type: boolean gates: type: boolean description: >- Whether this cell decides the release (calibrated, own caller, current epoch). gateReason: type: string result: type: string enum: - passed - failed - flaky - error - drift - not_run attempts: type: array items: type: object properties: runId: type: string description: >- The attempt's id: GET /v1/agents/{agentId}/attempts/{attemptId}. Named runId in schemaVersion 1. attempt: type: integer result: type: - string - 'null' enum: - passed - failed - error - drift - null required: - runId - attempt - result description: Each attempt's own result, in order. k: type: integer description: Attempts that passed. 'n': type: integer description: Attempts scored. change: type: - string - 'null' enum: - fixed - still_failing - newly_broken - still_passing - not_comparable - null reason: type: - string - 'null' criteria: type: array items: type: object properties: criterionId: type: string text: type: string met: type: - boolean - 'null' decidedBy: type: object properties: kind: type: string model: type: - string - 'null' answers: type: array items: {} required: - kind evidenceTurn: type: integer quote: type: string required: - criterionId - text - met - decidedBy noFixture: type: array items: type: string description: Tools that answered with the fixed 'not available' error. retellGrader: type: - object - 'null' properties: status: type: - string - 'null' explanation: type: - string - 'null' required: - status - explanation description: Retell's own grader on the last attempt. findings: type: array items: {} description: Problems Signal's own checks found on the last attempt's transcript. cost: type: object properties: retellUsd: type: number signalUsd: type: number required: - retellUsd - signalUsd required: - testId - testRevision - source - persona - tier - lane - expect - guard - gates - gateReason - result - attempts - k - 'n' - change - reason - criteria - noFixture - retellGrader - findings - cost ErrorType: type: string enum: - invalid_request_error - authentication_error - budget_error - permission_error - not_found_error - conflict_error - rate_limit_error - api_error securitySchemes: bearerAuth: type: http scheme: bearer description: An API key (`sig_...`), or a session token from the web app. ```` This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.