← Receivers

agent:run

agent family

mcp

Invoke a bound actor (optionally skill-constrained) for one bounded turn via channels; returns { text }

Effect
ask
Awaits an outcome — the call returns the response below.
Caller
manage_workflows
The class of authority the caller must already hold, decided from the attested context with no round trip.
Cost
variable
What one call costs, so you can budget before acting.
Reversible
no
This cannot be undone. Dry-run it first where the contract allows.
Idempotent
no
Not safe to blind-retry — dedupe on the envelope idempotencyKey.

Send it with your agent

One click hands your coding agent a prompt that registers the substrate, reads this contract, and makes the call. Launch opens the app; the others copy the prompt.

Claude Code
Codex
Cursor
Gemini CLI
Claude Desktop
ChatGPT
curl -X POST https://one.ie/api/ask/agent:run \
  -H "Authorization: Bearer $ONE_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"data": { "actorId": "one:support", "skill": "support", "instructions": "Reply kindly." }}'

The key is never in a link. npx -y @oneie/cli login writes it to ~/.config/oneie/key on your machine.

Request

Validated before dispatch — an invalid payload is refused with the fix, never half-applied.

  • actorId string optional
  • skill string optional
  • instructions string optional
  • model string optional Run this turn on a named model (an OpenRouter id, e.g. 'moonshotai/kimi-k3') instead of the actor's own. VETTED server-side against the workspace's enabledModelIds — a claim outside that list degrades to the operator's configured model, never to the claim. Absent = the actor's model, unchanged.
  • tags string[] optional Bare routing words for THIS call — a shape ('react-component'), a department ('sales'), a bake run id. Never namespaced. Carried onto the AI Gateway log as metadata beside the call's tokens and cost (channels/src/gateway-meta.ts), so spend can be filtered by them; never part of the user turn. The door's own signalId rides beside them from ctx, never from this body.

Response

What comes back from the call.

  • ok boolean
  • text string
  • actorId string
  • error string
  • usage object What this run cost, one-router card 11. Absent only when channels reported no usage.

Traffic

Every call to agent:run, counted where it is dispatched — over HTTP or in-process alike. Aggregate only — no actor, no payload, no workspace.

Counting…

Wiring

Every place in the open source that names agent:run, and the file that answers it. Read from the tree at build time — a receiver is reached by NAME through one door, so there is no import edge to follow and a grep is the honest shape of the question. Structure, not volume — the count is in Traffic above.

Called from

UI 3
  • one.ie/web/src/components/agents/AgentRun.tsx:29
  • one.ie/web/src/components/in/ParticipantPicker.tsx:130
  • one.ie/web/src/components/views/RowDoors.tsx:32
API route 9
  • one.ie/web/src/lib/ehc/as-story.ts:105
  • one.ie/web/src/lib/ehc/cases/cyp-001.ts:537
  • one.ie/web/src/lib/free-models.ts:4
  • one.ie/web/src/lib/in/assign.ts:30
  • one.ie/web/src/lib/resolvers/company.ts:20
  • one.ie/web/src/lib/resolvers/scheduling.ts:4
  • one.ie/web/src/lib/resolvers/tasks.ts:2022
  • one.ie/web/src/lib/step-ceiling.ts:17
  • one.ie/web/src/pages/api/chat.ts:1442
Agent runtime 2
  • channels/src/gateway-meta.ts:58
  • channels/src/tenant-endpoint.ts:13
Workflow 8
  • one.ie/web/src/components/workflows/StepPalette.tsx:36
  • one.ie/web/src/lib/factory/board-read.ts:335
  • one.ie/web/src/lib/factory/board.ts:57
  • one.ie/web/src/lib/factory/flow.ts:92
  • one.ie/web/src/lib/workflow-actor.ts:2
  • one.ie/web/src/lib/workflow-executor.ts:427
  • one.ie/web/src/lib/workflow-templates.ts:1186
  • one.ie/web/src/lib/workflows/campaign-ads-social.tql.ts:37
SDK 1
  • packages/sdk/src/schemas.ts:209
Tests 2
  • one.ie/web/src/lib/factory/gap-parity.test.ts:110
  • one.ie/web/src/lib/factory/wire-parity.test.ts:244
Docs 3
  • one.ie/web/src/content/docs/skills.md:26
  • one.ie/web/src/content/docs/workflows.md:40
  • one.ie/web/src/lib/CLAUDE.md:26
→

Answered by

agent:run one.ie/web/src/lib/workflow-receivers.ts:1892 Dispatched through POST /api/ask/agent:run, after the envelope validates the payload.
JSON Schema
{
  "$schema": "https://json-schema.org/draft/2020-12/schema",
  "type": "object",
  "properties": {
    "actorId": {
      "type": "string"
    },
    "skill": {
      "type": "string"
    },
    "instructions": {
      "type": "string"
    },
    "model": {
      "description": "Run this turn on a named model (an OpenRouter id, e.g. 'moonshotai/kimi-k3') instead of the actor's own. VETTED server-side against the workspace's enabledModelIds — a claim outside that list degrades to the operator's configured model, never to the claim. Absent = the actor's model, unchanged.",
      "type": "string",
      "maxLength": 64
    },
    "tags": {
      "description": "Bare routing words for THIS call — a shape ('react-component'), a department ('sales'), a bake run id. Never namespaced. Carried onto the AI Gateway log as metadata beside the call's tokens and cost (channels/src/gateway-meta.ts), so spend can be filtered by them; never part of the user turn. The door's own signalId rides beside them from ctx, never from this body.",
      "maxItems": 12,
      "type": "array",
      "items": {
        "type": "string",
        "minLength": 1,
        "maxLength": 64
      }
    }
  },
  "additionalProperties": {}
}
{
  "$schema": "https://json-schema.org/draft/2020-12/schema",
  "type": "object",
  "properties": {
    "ok": {
      "type": "boolean"
    },
    "text": {
      "type": "string"
    },
    "actorId": {
      "type": "string"
    },
    "error": {
      "type": "string"
    },
    "usage": {
      "description": "What this run cost, one-router card 11. Absent only when channels reported no usage.",
      "type": "object",
      "properties": {
        "model": {
          "type": "string"
        },
        "tokensIn": {
          "type": "number"
        },
        "tokensOut": {
          "type": "number",
          "description": "Provider-reported output tokens, INCLUDING reasoning — the part the answer text never shows"
        },
        "reasoningTokens": {
          "type": "number"
        },
        "credits": {
          "description": "What the workspace was charged: upstream cost + margin, whole credits",
          "type": "number"
        },
        "costCredits": {
          "description": "Upstream cost alone, priced per model from the live OpenRouter blob",
          "type": "number"
        },
        "refused": {
          "description": "A gate that would have refused this spend (402 | floored | budget_exhausted | float_locked). The row was still written: the tokens were already spent",
          "type": "string"
        },
        "unmetered": {
          "description": "No burn row was written — say so, never read it as free",
          "type": "boolean",
          "const": true
        }
      },
      "required": [
        "model",
        "tokensIn",
        "tokensOut"
      ]
    }
  },
  "required": [
    "ok"
  ]
}