Tale · Schema
RunSummary
One run as a listing reports it — identity, scope, status and timing. The input, output, trace, effects and checkpoints of the full `Run` are inlined only when `include` names them; the single read answers the full row.
CompanyAIAgentsAutomationKnowledge ManagementMCPOpen SourceSelf-Hosted
Properties
| Name | Type | Description |
|---|---|---|
| id | string | The run id — the name the full read uses; `runId` carries the same value |
| runId | string | The run id again, under the name the start answered — one value under both names |
| name | string | The automation the run was started as. Runs outlive their automation: after `DELETE /api/v1/automations/{name}` the rows stay listed and readable by id under this name, which `GET /api/v1/automations/ |
| version | integer | |
| projectId | stringnull | URL project for a project run; null for an organization run |
| status | string | |
| mode | string | |
| startedBy | string | What started the run — one of four forms: `api-key: |
| startedVia | string | Which kind of trigger started a `trigger: |
| detail | string | The failure or wait reason, when the run has one. While `waiting` it names the park: `approval: |
| failureCode | string | Why a `failed` run failed — present only with `status: "failed"` on a run that failed on a build that records it. See `Run.failureCode` for the vocabulary and which codes are worth a retry. |
| waitingFor | string | Present only while `status` is `waiting`: what the run is parked on. `approval` — a person’s decision on a gate; `ask` — a question a person has to answer; `in_doubt` — a write the run was making when |
| legacyQuarantine | object | |
| resumeCount | integer | How often the run moved to another server: a server that was being updated or restarted handed it on, or another one took it over after its own stopped responding. Steps it had finished never run agai |
| lastResume | object | Why and when the run last moved to another server; absent while it never did. `shutdown` — its server was being updated or restarted and handed it on; `lease_expired` — its server stopped responding a |
| stalled | boolean | True while a `running` run waits for a server to take it over after its own stopped (the app reads it as “Interrupted — resuming”): nothing is working on it right now, and another server picks it up w |
| startedAt | integer | Epoch milliseconds |
| finishedAt | integer | Epoch ms; absent while the run is not terminal |
| input | object | With `include=input` |
| output | object | With `include=output` |
| trace | object | With `include=trace` |
| effects | object | With `include=effects` |
| checkpoints | object | With `include=checkpoints` |
JSON Schema
{
"$schema": "https://json-schema.org/draft/2020-12/schema",
"$id": "https://raw.githubusercontent.com/api-evangelist/tale-dev/main/json-schema/tale-dev-run-summary-schema.json",
"title": "RunSummary",
"description": "One run as a listing reports it — identity, scope, status and timing. The input, output, trace, effects and checkpoints of the full `Run` are inlined only when `include` names them; the single read answers the full row.",
"x-generated": "2026-10-09",
"x-method": "derived",
"x-generator": "derive-json-schema.py",
"x-source": "openapi/tale-dev-openapi.yml#/components/schemas/RunSummary",
"type": "object",
"required": [
"id",
"runId",
"name",
"version",
"projectId",
"status",
"mode",
"startedBy",
"startedAt"
],
"properties": {
"id": {
"type": "string",
"description": "The run id — the name the full read uses; `runId` carries the same value"
},
"runId": {
"type": "string",
"description": "The run id again, under the name the start answered — one value under both names"
},
"name": {
"type": "string",
"description": "The automation the run was started as. Runs outlive their automation: after `DELETE /api/v1/automations/{name}` the rows stay listed and readable by id under this name, which `GET /api/v1/automations/{name}` then answers 404 for."
},
"version": {
"type": "integer"
},
"projectId": {
"type": [
"string",
"null"
],
"description": "URL project for a project run; null for an organization run"
},
"status": {
"type": "string",
"enum": [
"queued",
"running",
"waiting",
"quarantined",
"success",
"failed",
"cancelled"
]
},
"mode": {
"type": "string",
"enum": [
"mock",
"live"
]
},
"startedBy": {
"type": "string",
"description": "What started the run — one of four forms: `api-key:<userId>` (a start through this API or the MCP endpoint), `user:<userId>` (a start from the product), `trigger:<triggerId>` (a schedule, webhook or event — the id the trigger read answers), or a bare user id on runs the in-product builder recorded before it prefixed them. Split on the first `:`; treat a value without one as a user id."
},
"startedVia": {
"type": "string",
"enum": [
"schedule",
"webhook",
"event"
],
"description": "Which kind of trigger started a `trigger:<id>` run — read off the run’s own input, so it stays true after the binding changes kind. Absent on a run a person or an API key started."
},
"detail": {
"type": "string",
"description": "The failure or wait reason, when the run has one. While `waiting` it names the park: `approval:<approvalId>`, `agent:<nodeId>`, `room:<nodeId>`, `repeat:<nodeId>` or `in_doubt:<nodeId>` — `waitingFor` is the field to branch on; when `failed`, the failure sentence, and `failureCode` the stable cause — the sentence is not contractual."
},
"failureCode": {
"type": "string",
"enum": [
"node_error",
"connector_error",
"llm_output_invalid",
"approval_rejected",
"execution_limit",
"automation_deleted",
"engine_incompatible",
"effect_in_doubt",
"missing_api_key",
"credit_exhausted",
"model_not_entitled",
"auth_error",
"model_not_found",
"unsupported_parameter",
"output_cap_too_high",
"token_limit",
"rate_limited",
"content_filter",
"context_length",
"provider_unreachable",
"provider_error",
"budget_exceeded",
"harness_error",
"turn_crashed",
"session_gone",
"start_failed",
"harvest_failed",
"resume_failed",
"deadline",
"ask_expired"
],
"description": "Why a `failed` run failed — present only with `status: \"failed\"` on a run that failed on a build that records it. See `Run.failureCode` for the vocabulary and which codes are worth a retry."
},
"waitingFor": {
"type": "string",
"enum": [
"approval",
"ask",
"in_doubt",
"agent",
"room",
"repeat"
],
"description": "Present only while `status` is `waiting`: what the run is parked on. `approval` — a person’s decision on a gate; `ask` — a question a person has to answer; `in_doubt` — a write the run was making when its server stopped may or may not have reached its service, and a person must decide how to continue (in the app: run it again, skip it, or fail the run); `agent` — an agent turn still running, no one to page; `room` — an agent turn whose start waits for sandbox room, no one to page; `repeat` — a node polling until its `repeatUntil` condition holds, no one to page. \"Runs that need a human\" is `waitingFor` in (`approval`, `ask`, `in_doubt`), never `status=waiting` alone."
},
"legacyQuarantine": {
"$ref": "#/$defs/LegacyRunQuarantine"
},
"resumeCount": {
"type": "integer",
"minimum": 0,
"description": "How often the run moved to another server: a server that was being updated or restarted handed it on, or another one took it over after its own stopped responding. Steps it had finished never run again. `Run` always carries it; `RunSummary` omits it while it is 0."
},
"lastResume": {
"type": "object",
"required": [
"reason",
"at"
],
"additionalProperties": false,
"description": "Why and when the run last moved to another server; absent while it never did. `shutdown` — its server was being updated or restarted and handed it on; `lease_expired` — its server stopped responding and another one took it over.",
"properties": {
"reason": {
"type": "string",
"enum": [
"shutdown",
"lease_expired"
]
},
"at": {
"type": "integer",
"minimum": 0,
"description": "Epoch milliseconds"
}
}
},
"stalled": {
"type": "boolean",
"description": "True while a `running` run waits for a server to take it over after its own stopped (the app reads it as “Interrupted — resuming”): nothing is working on it right now, and another server picks it up within about a minute and a half. `Run` always carries it; `RunSummary` carries it only while true."
},
"startedAt": {
"type": "integer",
"minimum": 0,
"description": "Epoch milliseconds"
},
"finishedAt": {
"type": "integer",
"minimum": 0,
"description": "Epoch ms; absent while the run is not terminal"
},
"input": {
"description": "With `include=input`"
},
"output": {
"description": "With `include=output`"
},
"trace": {
"description": "With `include=trace`"
},
"effects": {
"description": "With `include=effects`"
},
"checkpoints": {
"description": "With `include=checkpoints`"
}
},
"$defs": {
"LegacyRunQuarantine": {
"type": "object",
"additionalProperties": false,
"description": "A legacy execution whose external effects or termination cannot be proven. Present on held runs. A recorded stop request does not resolve the hold or establish that work stopped.",
"required": [
"reason",
"observedAt",
"claimEpoch",
"priorStatus",
"resolution"
],
"properties": {
"reason": {
"type": "string",
"enum": [
"legacy_execution_unproven"
]
},
"observedAt": {
"type": "integer",
"minimum": 0,
"maximum": 9007199254740991
},
"claimEpoch": {
"type": "integer",
"minimum": 0,
"maximum": 9007199254740991
},
"priorStatus": {
"type": "string",
"enum": [
"queued",
"running",
"waiting"
]
},
"resolution": {
"type": [
"object",
"null"
],
"additionalProperties": false,
"required": [
"action",
"actor",
"at"
],
"properties": {
"action": {
"type": "string",
"enum": [
"stop"
]
},
"actor": {
"type": "string",
"minLength": 1,
"maxLength": 200
},
"at": {
"type": "integer",
"minimum": 0,
"maximum": 9007199254740991
}
}
}
}
}
}
}
Work with this as data
Every JSON Schema here is available over the APIs.io API and to AI agents over MCP.
MCP server
One button, every client — Claude, Cursor, VS Code and the rest.
https://apis.io/mcp
Tools for schemas
4 MCP tools reach this
find_json_schemasBrowse and filter every JSON Schema in the catalog.apis_io_searchSTART HERE — APIs, providers and tags for one query, each with its total.resolveTurn a domain, URL or GitHub org into the provider it belongs to.find_cohortsEvery scored population of providers in the catalog.
Call it yourself
curl for this page
This JSON Schema
curl "https://apis.io/api/v1/json-schemas/tale-dev-run-summary"
All schemas
curl "https://apis.io/api/v1/json-schemas?limit=25"
Discovery needs no key. Ratings and market analysis are Pro.
Get an API key
Free tier, no form to fill in. Signing in shares your email address with us — we store it to create your key and to recognise you if you sign in with another provider. See our Privacy Policy and Terms.
A second provider on the same verified email joins the account you already have.