> ## 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 call > One call with its analysis, who called and why, and the open issues it is evidence for. The transcript is in the call's trace. ## OpenAPI ````yaml /api-reference/openapi.json get /v1/agents/{agentId}/calls/{callId} 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}/calls/{callId}: get: tags: - Calls summary: Get a call description: >- One call with its analysis, who called and why, and the open issues it is evidence for. The transcript is in the call's trace. operationId: getCall 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 description: The call's id in Signal (not the platform's call id). required: true description: The call's id in Signal (not the platform's call id). name: callId in: path responses: '200': description: The call. content: application/json: schema: $ref: '#/components/schemas/Call' '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/calls/0b9e8d7c-6f5a-4b3c-9d2e-1f0a9b8c7d6e" \ -H "Authorization: Bearer $SIGNAL_API_KEY" components: schemas: Call: allOf: - $ref: '#/components/schemas/CallSummary' - type: object properties: analysis: type: - object - 'null' properties: intentConfidence: type: number description: 0 to 1. objectiveFulfilled: type: - boolean - 'null' description: >- Whether the caller got what they came for. Null when it could not be judged. fulfillmentConfidence: type: number description: 0 to 1. reasoning: type: - string - 'null' description: Why the outcome was read as it was. source: type: string description: >- Which reading produced this analysis. For information only; do not branch on it. analyzedAt: type: string format: date-time example: '2026-10-02T14:03:07.120Z' frictions: type: array items: type: object properties: type: type: string description: >- A signal code, or a kind the analysis named. `checklist.*` types are readings of what was said, not measurements. label: type: string klass: type: string enum: - technical - flow - content - behaviour description: >- Which kind of fix it points at: `technical` (an engineer), `flow` (whoever owns the conversation design), `content` (whoever owns prompts and knowledge), or `behaviour` (something callers do that is worth knowing but is not a defect). confidence: type: number description: 0 to 1. location: type: object properties: kind: type: string enum: - node - edge - tool - global description: >- A step of the flow, a transition between two steps, a tool, or the whole call. id: type: string description: >- The step id, `from->to` for a transition, the tool name, or `global`. label: type: string description: How the place reads to a person. required: - kind - id - label cause: type: - object - 'null' properties: code: type: string description: Stable key for the cause. label: type: string description: How the cause reads to a person. detail: type: - string - 'null' description: What in the trace says so. required: - code - label - detail description: >- Why it happened, read from what the trace shows around it. Null when no cause was read. turnIndex: type: - integer - 'null' description: >- The turn it is about, an index into the trace's `turns`. quote: type: - string - 'null' explanation: type: string required: - type - label - klass - confidence - location - cause - turnIndex - quote - explanation description: >- What the analysis noticed on the call, above `info` severity. Symptoms, not diagnosed issues; `issues` lists the issues this call is evidence for. required: - intentConfidence - objectiveFulfilled - fulfillmentConfidence - reasoning - source - analyzedAt - frictions description: The call's analysis. Null until the call is analysed. alsoAsked: type: array items: type: object properties: slug: type: string label: type: string coverage: type: - string - 'null' enum: - designed - out_of_scope - no_instructions - not_established - null description: >- Whether the agent was built for this use case: `designed`, `out_of_scope` (it says it does not handle it), `no_instructions` (nothing in its design covers it), or `not_established` (it could not be settled). Null when not checked. required: - slug - label - coverage description: Use cases the caller asked for besides `useCase`. traits: type: array items: type: object properties: slug: type: string label: type: string trait: type: string enum: - customer_type - demeanour - speech - surroundings - language required: - slug - label - trait description: >- The caller's traits as classified: who they are to the business, demeanour, speech, surroundings, language. coverage: type: object properties: useCase: type: - string - 'null' enum: - designed - out_of_scope - no_instructions - not_established - null description: >- Whether the agent was built for this use case: `designed`, `out_of_scope` (it says it does not handle it), `no_instructions` (nothing in its design covers it), or `not_established` (it could not be settled). Null when not checked. subUseCase: type: - string - 'null' enum: - designed - out_of_scope - no_instructions - not_established - null description: >- Whether the agent was built for this use case: `designed`, `out_of_scope` (it says it does not handle it), `no_instructions` (nothing in its design covers it), or `not_established` (it could not be settled). Null when not checked. required: - useCase - subUseCase description: Whether the agent was built for what this caller wanted. audio: type: - object - 'null' properties: noiseDb: type: number description: >- Median level of the caller's channel outside their turns, in dBFS. speechDb: type: - number - 'null' description: >- Loud end of the caller's channel inside their turns, in dBFS. snrDb: type: - number - 'null' description: '`speechDb - noiseDb`.' required: - noiseDb - speechDb - snrDb description: >- What the recording measured on the caller's side. Null when it was not read. issues: type: array items: $ref: '#/components/schemas/CallIssue' description: >- Open issues this call is evidence for, highest priority first. Resolved and withdrawn issues are left out. required: - analysis - alsoAsked - traits - coverage - audio - issues example: object: call id: 0b9e8d7c-6f5a-4b3c-9d2e-1f0a9b8c7d6e providerCallId: call_5e0c2b7a91d34f6e8a1b version: 8 startedAt: '2026-10-01T15:42:05.000Z' endedAt: '2026-10-01T15:44:51.000Z' durationMs: 166000 direction: inbound disconnectionReason: user_hangup outcome: abandoned objective: Book a cleaning for next week intent: slug: book_appointment label: Book appointment useCase: slug: scheduling label: Scheduling subUseCase: slug: new_booking label: New booking callerPersona: slug: hurried-returning-patient label: Hurried returning patient caller: kind: repeat nth: 2 of: 3 counts: turns: 24 callerTurns: 12 agentTurns: 12 toolFailures: 1 frictions: 2 signalCodes: - tool_timeout - dead_air analysis: intentConfidence: 0.94 objectiveFulfilled: false fulfillmentConfidence: 0.81 reasoning: >- The caller asked for a cleaning next week; the availability lookup did not answer and the caller hung up before a time was offered. source: llm analyzedAt: '2026-10-01T15:47:30.000Z' frictions: [] alsoAsked: - slug: insurance_question label: Insurance question coverage: designed traits: - slug: in_a_hurry label: In a hurry trait: demeanour coverage: useCase: designed subUseCase: designed audio: noiseDb: -58.2 speechDb: -21.4 snrDb: 36.8 issues: - object: issue id: a2b3c4d5-e6f7-4a8b-9c0d-1e2f3a4b5c6d key: tool_timeout:check_availability title: The agent did not reply in Book appointment status: discovered klass: technical Error: type: object properties: error: type: object properties: type: $ref: '#/components/schemas/ErrorType' code: type: string example: agent_not_found description: The specific reason; stable within v1. message: type: string example: Agent not found. param: type: - string - 'null' description: The request field at fault, for a validation error. requestId: type: - string - 'null' example: req_8c1f0e2b4a6d9e3f7a5c1b0d quota: type: object properties: metric: type: string example: test_calls description: analysed_calls, agents, copilot_messages or test_calls. limit: type: integer used: type: integer plan: type: string example: free_org description: The workspace's plan slug. upgradePath: type: string example: /billing description: Where in the web app the plan is changed. required: - metric - limit - used - plan - upgradePath description: >- Present when code is quota_exceeded: the plan limit the request would pass. required: - type - code - message - param - requestId required: - error example: error: type: not_found_error code: agent_not_found message: Agent not found. param: null requestId: req_8c1f0e2b4a6d9e3f7a5c1b0d CallSummary: type: object properties: object: type: string enum: - call id: type: string format: uuid pattern: >- ^[0-9a-fA-F]{8}-[0-9a-fA-F]{4}-[0-9a-fA-F]{4}-[0-9a-fA-F]{4}-[0-9a-fA-F]{12}$ example: 3f1c2d4e-5a6b-4c7d-8e9f-0a1b2c3d4e5f description: The call's id in Signal. providerCallId: type: string description: >- The call's id on the voice platform. Unique within an agent, not across agents. version: type: - integer - 'null' description: The agent version that took the call. startedAt: type: - string - 'null' format: date-time example: '2026-10-02T14:03:07.120Z' endedAt: type: - string - 'null' format: date-time example: '2026-10-02T14:03:07.120Z' durationMs: type: - integer - 'null' direction: type: - string - 'null' description: '`inbound` or `outbound`, as the platform reports it.' disconnectionReason: type: - string - 'null' description: >- Why the call ended, in the platform's words (`user_hangup`, `agent_hangup`, ...). outcome: type: - string - 'null' enum: - resolved - unresolved - transferred - abandoned - no_contact - null description: How the call ended for the caller. Null until the call is analysed. objective: type: - string - 'null' description: >- One line, in the caller's terms, of what they were trying to do. Null until analysed. intent: type: - object - 'null' properties: slug: type: string label: type: string required: - slug - label description: The intent the analysis gave the call. Null until analysed. useCase: type: - object - 'null' properties: slug: type: string label: type: string required: - slug - label description: >- What the caller wanted, from the use case classification. Null when not classified. subUseCase: type: - object - 'null' properties: slug: type: string label: type: string required: - slug - label description: The narrower use case under `useCase`, when there is one. callerPersona: type: - object - 'null' properties: slug: type: string label: type: string required: - slug - label description: The kind of caller the call was read as. Null when not classified. caller: type: object properties: kind: type: string enum: - test - automated - silent - repeat - new description: >- Who called, from facts alone. `test`: a line the agent's own team tests it from; such a call counts in no issue. `automated`: another company's AI agent. `silent`: nobody on the caller's side spoke. `repeat`: a number that called this agent more than once. `new`: none of these. nth: type: - integer - 'null' description: >- For `repeat`: which of that number's calls this is, oldest first. of: type: - integer - 'null' description: 'For `repeat`: how many calls that number made to this agent.' required: - kind - nth - of description: >- Who called, from facts alone. The phone number itself is never returned. counts: type: object properties: turns: type: - integer - 'null' description: Turns in the transcript. Null when the call has no trace. callerTurns: type: - integer - 'null' agentTurns: type: - integer - 'null' toolFailures: type: - integer - 'null' description: >- Tool calls that failed. A tool that ended the call is not a failure. frictions: type: - integer - 'null' description: >- Points of friction the analysis recorded (see the call's `analysis.frictions`). Null until analysed. required: - turns - callerTurns - agentTurns - toolFailures - frictions signalCodes: type: - array - 'null' items: type: string description: >- Codes of the signals detected on the call, sorted. A signal is a symptom noticed on a call, not a diagnosed issue; what to fix is on the issues. Null until analysed, empty when nothing was noticed. required: - object - id - providerCallId - version - startedAt - endedAt - durationMs - direction - disconnectionReason - outcome - objective - intent - useCase - subUseCase - callerPersona - caller - counts - signalCodes example: object: call id: 0b9e8d7c-6f5a-4b3c-9d2e-1f0a9b8c7d6e providerCallId: call_5e0c2b7a91d34f6e8a1b version: 8 startedAt: '2026-10-01T15:42:05.000Z' endedAt: '2026-10-01T15:44:51.000Z' durationMs: 166000 direction: inbound disconnectionReason: user_hangup outcome: abandoned objective: Book a cleaning for next week intent: slug: book_appointment label: Book appointment useCase: slug: scheduling label: Scheduling subUseCase: slug: new_booking label: New booking callerPersona: slug: hurried-returning-patient label: Hurried returning patient caller: kind: repeat nth: 2 of: 3 counts: turns: 24 callerTurns: 12 agentTurns: 12 toolFailures: 1 frictions: 2 signalCodes: - tool_timeout - dead_air CallIssue: type: object properties: object: type: string enum: - issue id: 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 key: type: string description: The issue's stable key; tests take an issue by it. title: type: string status: type: string description: >- `discovered`, `investigating`, `fix_deployed`, `measuring` or `regressed`. klass: type: string enum: - technical - flow - content - behaviour description: >- Which kind of fix it points at: `technical` (an engineer), `flow` (whoever owns the conversation design), `content` (whoever owns prompts and knowledge), or `behaviour` (something callers do that is worth knowing but is not a defect). required: - object - id - key - title - status - klass 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.