> ## Documentation Index > Fetch the complete documentation index at: https://docs.corti.ai/llms.txt > Use this file to discover all available pages before exploring further. # Get an agent's usage > Returns invocation metrics for the agent over the half-open `[from, to)` time range (UTC), bucketed at the requested `granularity`. The response echoes the resolved range and granularity, a `totals` summary across the whole range, and one `buckets` entry per period that had activity (the array is empty when there was none). When `from`/`to` are omitted, the range defaults to the last 30 days. ## OpenAPI ````yaml /agentic/auto-generated-openapi-v2.yml get /v2/agentic/agents/{agentId}/usage openapi: 3.1.0 info: title: Corti Agent API v2 version: 2.0.0 summary: Manage agents and converse with them over the A2A v1.0 protocol. description: > Version 2 of the Corti Agents REST API. ## What's new in v2 - **Unified `connectors`** — a single, flat, discriminated array replaces the v1 split between `experts`, `mcpServers`, and sub-agents. The envelope is extensible: new connector kinds (`a2a`, `openapi`, `custom`) slot in behind the `type` discriminator without breaking changes. - **Clean CRUD verbs** — `POST` creates, `GET` fetches/lists, `PATCH` performs a *true* partial update, `DELETE` removes. v1's PATCH was effectively a PUT. - **First-class metadata** — `visibility`, `model`, `lifecycle` (now a body field, not a query param), and free-form `labels`. ## A2A v1.0 only The conversational surface speaks **A2A protocol version `1.0` exclusively** (both the `JSONRPC` and `HTTP+JSON` bindings). The deprecated v0.3 binding from v1 is intentionally not carried forward. Per A2A §3.6, clients MUST send the `A2A-Version` header (`Major.Minor`, e.g. `1.0`) on every request to the `/a2a/*` surface, or supply it as the `A2A-Version` query parameter. This surface implements `1.0` only; an absent header is treated as `1.0`. Patch versions MUST NOT be sent. The server echoes the negotiated version in the `A2A-Version` response header. ## Partial updates (PATCH) `PATCH` uses **JSON Merge Patch** semantics (RFC 7386), served under the `application/merge-patch+json` media type: - **Omit a field** → left unchanged. - **`null`** → cleared / reset to its default. - **Arrays** (e.g. `connectors`) → *replaced wholesale*, never merged. To mutate a single connector without replacing the whole array, use the `/connectors` sub-resource endpoints. ## Streaming (Server-Sent Events) The streaming A2A surfaces — `message/stream`, `tasks/subscribe`, and the streaming JSON-RPC methods — respond with `text/event-stream`. The `text/event-stream` media type carries a sequence of Server-Sent Events; the attached schema describes a single event frame. Every event's `data` field is a string carrying a JSON document (`contentMediaType: application/json`); `contentSchema` declares the parsed payload (`StreamResponse` for the `HTTP+JSON` binding, `JSONRPCResponse` for the `JSONRPC` binding). The optional `id` field carries the SSE event id used for resumption via the `Last-Event-ID` request header. contact: name: Corti API Support url: https://corti.ai email: support@corti.ai license: name: Proprietary url: https://corti.ai/terms termsOfService: https://corti.ai/terms servers: - url: https://api.{env}.corti.app description: Corti regional API gateway variables: env: default: eu enum: - eu - us description: Deployment region. security: - bearerAuth: [] tenantHeader: [] tags: - name: Agents description: Create, read, update, and delete agents. - name: Agent Card description: A2A-compliant agent discovery cards. - name: A2A description: Converse with an agent over the A2A v1.0 protocol. - name: Connectors description: Manage the connectors attached to an agent. - name: Usage description: Bucketed invocation history for an agent. - name: Contexts description: Conversational contexts and their tasks. - name: Registry description: Discoverable, pre-built connectors offered by the platform. - name: Artifacts description: Retrieve artifacts produced by tasks. - name: Feedback description: Collect human or automated feedback on tasks and messages. paths: /v2/agentic/agents/{agentId}/usage: parameters: - $ref: '#/components/parameters/AgentId' get: tags: - Usage summary: Get an agent's usage description: | Returns invocation metrics for the agent over the half-open `[from, to)` time range (UTC), bucketed at the requested `granularity`. The response echoes the resolved range and granularity, a `totals` summary across the whole range, and one `buckets` entry per period that had activity (the array is empty when there was none). When `from`/`to` are omitted, the range defaults to the last 30 days. operationId: UsageGet parameters: - name: from in: query description: | Inclusive start of the range, as an RFC 3339 timestamp (UTC). Defaults to 30 days before `to`. Must not be after `to`. schema: type: string format: date-time example: '2026-05-19T00:00:00Z' - name: to in: query description: | Exclusive end of the range, as an RFC 3339 timestamp (UTC). Defaults to the current time. schema: type: string format: date-time example: '2026-05-20T00:00:00Z' - name: granularity in: query description: Size of each reporting bucket. Defaults to `day`. schema: $ref: '#/components/schemas/UsageGranularity' example: day responses: '200': description: The agent's bucketed usage over the requested range. content: application/json: schema: $ref: '#/components/schemas/UsageReportResponse' '400': $ref: '#/components/responses/BadRequest' '401': $ref: '#/components/responses/Unauthorized' '404': $ref: '#/components/responses/NotFound' components: parameters: AgentId: name: agentId in: path required: true description: Agent identifier (prefixed UUIDv7). schema: $ref: '#/components/schemas/CommonAgentIDValue' schemas: UsageGranularity: type: string description: >- The size of each usage reporting bucket. Only `day` is currently honored; `minute`, `hour`, and `week` are accepted but produce daily buckets (the server always returns `day`). enum: - minute - hour - day - week default: day UsageReportResponse: type: object description: An agent's bucketed usage over a date range, with range-wide totals. required: - granularity - from - to - totals - buckets properties: granularity: $ref: '#/components/schemas/UsageGranularity' from: type: string format: date-time description: Resolved inclusive start of the range (UTC). examples: - '2026-05-19T00:00:00Z' to: type: string format: date-time description: Resolved exclusive end of the range (UTC). examples: - '2026-05-21T00:00:00Z' totals: $ref: '#/components/schemas/UsageMetrics' description: Aggregate metrics across the whole range. buckets: type: array description: One entry per period with activity, ordered oldest first. items: $ref: '#/components/schemas/UsageBucket' examples: - granularity: day from: '2026-05-19T00:00:00Z' to: '2026-05-21T00:00:00Z' totals: invocations: 15 uniqueContexts: 6 buckets: - periodStart: '2026-05-19T00:00:00Z' periodEnd: '2026-05-20T00:00:00Z' invocations: 12 uniqueContexts: 5 - periodStart: '2026-05-20T00:00:00Z' periodEnd: '2026-05-21T00:00:00Z' invocations: 3 uniqueContexts: 2 CommonAgentIDValue: type: string format: agent-id description: >- Agent identifier. Accepts `agt.` or a bare UUIDv7 on input; always returned prefixed. pattern: ^(agt\.)?[0-9a-f]{8}-[0-9a-f]{4}-[0-9a-f]{4}-[0-9a-f]{4}-[0-9a-f]{12}$ examples: - agt.0192f4c8-2c5a-7b3e-9f1a-3c8d6e2b7a40 UsageMetrics: type: object description: Invocation metrics for a single period. required: - invocations - uniqueContexts properties: invocations: type: integer format: int64 minimum: 0 description: Number of agent invocations in the period. examples: - 12 uniqueContexts: type: integer format: int64 minimum: 0 description: Number of distinct contexts invoked in the period. examples: - 5 UsageBucket: description: Usage metrics for a single time bucket. allOf: - type: object required: - periodStart - periodEnd properties: periodStart: type: string format: date-time description: Inclusive start of the bucket (UTC). examples: - '2026-05-19T00:00:00Z' periodEnd: type: string format: date-time description: Exclusive end of the bucket (UTC). examples: - '2026-05-20T00:00:00Z' - $ref: '#/components/schemas/UsageMetrics' CommonErrorResponse: type: object description: > Corti management-plane error envelope, used by all non-A2A endpoints. - **Standard** — when the error chain contains at least one `PublicError`, `code` and `message` come from the outermost `PublicError` and `details` is merged across the whole chain (outer values take precedence). - **Fallback** — when the chain contains no `PublicError`, the response is a generic `500` carrying a `requestId` for support reference. - **Validation** — a single `PublicError` whose `details.validationErrors` lists the offending fields. Field names use camelCase on the wire (e.g. `requestId`, `howToFix`). The free-form `details` object may carry arbitrary caller-defined keys. Rate limiting (HTTP 429) is not yet implemented; the server does not emit a 429 response. required: - error properties: error: type: object required: - code - message properties: code: type: string description: Stable, machine-readable, SCREAMING_SNAKE_CASE error code. examples: - ASSIGNMENT_CONFLICT - VALIDATION_FAILED - INTERNAL_ERROR message: type: string description: Human-readable explanation. howToFix: type: string description: Optional guidance for the caller to resolve the error. details: type: object description: | Structured context, merged from every `PublicError` in the chain (outer values win). Omitted on the generic fallback response. additionalProperties: true properties: validationErrors: type: array description: Present when `code` is `VALIDATION_FAILED`. items: type: object required: - field - reason properties: field: type: string description: The field that failed validation. reason: type: string description: Why the field failed validation. requestId: type: string description: > Correlation ID from request middleware. Included only on the generic `500` fallback so consumers can quote it in support requests. description: The error object with code, message, and optional details. examples: - error: code: ASSIGNMENT_CONFLICT message: could not complete assignment details: assignment_id: asg_99 expert_id: exp_42 - error: code: VALIDATION_FAILED message: validation failed details: validationErrors: - field: email reason: invalid format - field: name reason: required - error: code: INTERNAL_ERROR message: internal server error requestId: req_abc123 responses: BadRequest: description: The request was malformed. content: application/json: schema: $ref: '#/components/schemas/CommonErrorResponse' Unauthorized: description: Missing or invalid credentials. content: application/json: schema: $ref: '#/components/schemas/CommonErrorResponse' NotFound: description: The resource does not exist or is not visible to the caller. content: application/json: schema: $ref: '#/components/schemas/CommonErrorResponse' securitySchemes: bearerAuth: type: http scheme: bearer bearerFormat: JWT description: OAuth 2.0 / OIDC bearer token. tenantHeader: type: apiKey in: header name: Tenant-Name description: The tenant the request operates within. ````