Publish a rule
Success status: 202 Accepted.
POST /v25-10-beta/agentic-logic/rules/{ruleId}/publish
Starts an asynchronous publish job for the rule's current draft (code rules) or template configuration (builder rules).
- notes (string, optional) — Notes stamped on the resulting version.
- title (string, optional, up to 255 characters) — Display title stamped on the resulting version. When omitted, a draft that came from the AI chat gets "Built from chat" and any other draft gets no title.
- idempotencyKey (string, optional) — Client-supplied key. Repeating a publish for the same rule and key returns the existing job handle instead of starting another job.
- autoAssist (boolean, optional) — Whether a failed job starts AI repair. Defaults to true. Send false to leave a failed job failed (see below). It is frozen into this job only; it is not stored on the rule.
The endpoint answers 202 Accepted with { "publishJobId": "..." }. Publishing runs a gate pipeline (static checks, tests, budget and safety gates, deployment) that takes seconds to minutes; poll the publish job until status is succeeded, failed, or cancelled. A successful publish makes the new version active immediately. A first publish moves a draft rule to active; a publish never resumes a paused rule.
- When a code rule's draft has no validation scenarios, the checks run one default scenario: an onsite request in USD from the US, with the rule's id as the placement.
- A failed publish of a code rule opens an AI repair run, unless one is already running. The repair loop rewrites the draft and re-runs the checks; when they pass, it publishes the rewritten source, unless the rewrite changed what the rule does. Follow it through activeRepairRun on Get a rule; the failed job itself stays failed.
With autoAssist: false, a failed job opens no repair run, so the platform never rewrites the rule's source or republishes it on its own. It does not stop a repair run that is already running; cancel does. A successful publish is unaffected.
A repeated idempotencyKey returns the job it first started, with the autoAssist value that job was started with. The autoAssist sent on the repeat is ignored.
Use the intelligems-access-token header. The token selects the organization; do not send orgId. The organization must have Agentic Logic enabled.
Rate Limit: 50 requests per 30s window, refilling 25 tokens every 30s.
Request
- Base URLhttps://api.intelligems.io
- AuthAPI key in header intelligems-access-token
- TypeScriptA function with typed inputs. Its response is checked against the schema.
Path parameters
Request body
Response
Accepted
Changes
No recorded changes to this operation across all 1 revision of this API.