All tools
ToolWrite

Run Research Thread

run_research_thread

Use only when the user wants a bounded deepen, verify, red-team, or refresh investigation of an existing Thread.

When to use it

Use only when the user wants a bounded deepen, verify, red-team, or refresh investigation of an existing Thread.

Typical workflow

List existing Threads first, inspect the case file, then add exact evidence, propose a revision, or start one bounded run and poll it to a terminal state.

What it returns

A runId, queued or running status, pollAfterMs, and a bounded run record; poll with get_research_thread_run.

Avoid using it when

Do not use merely to read the case file or when the user has not authorized metered model work.

Advanced input contract

Agent hosts use this JSON Schema to validate the exact input shape.

View full registry notes

Start a bounded, metered deepen, verify, red-team, or refresh run for one Research Thread. Completion may add exact evidence and update non-terminal tracking analysis, but cannot validate, reject, incorporate, apply a revision, or change the canonical thesis. Returns a runId for polling. Use when: Use only when the user wants a bounded deepen, verify, red-team, or refresh investigation of an existing Thread. Avoid when: Do not use merely to read the case file or when the user has not authorized metered model work.

View JSON Schema
{
  "type": "object",
  "properties": {
    "threadId": {
      "type": "string",
      "format": "uuid"
    },
    "mode": {
      "type": "string",
      "enum": [
        "deepen",
        "verify",
        "red_team",
        "refresh"
      ]
    },
    "prompt": {
      "type": "string",
      "maxLength": 4000
    },
    "idempotencyKey": {
      "type": "string",
      "minLength": 1,
      "maxLength": 500
    }
  },
  "required": [
    "threadId",
    "mode",
    "idempotencyKey"
  ],
  "additionalProperties": false
}

Useful next tools