canonical: https://jentic.com/apis/runframe.io/runframe

# Runframe V1 API

The Runframe V1 API runs incident response for engineering teams: it opens incidents against services, moves them through acknowledgement, status transitions, and escalation, and records postmortems once they are over. Responders can be paged by email address over Slack or email, the current on-call person for a team can be looked up, and escalation policies are readable per severity. A set of public status routes returns the status page summary, components, incident history, and maintenance windows without a key.

## For AI agents

Open, acknowledge, escalate, and resolve Runframe incidents, page responders over Slack or email, look up who is on call, and write postmortems.

## Scope

Does not collect metrics, ingest logs, or edit on-call schedules. Use for managing incidents, paging responders, and writing postmortems only.

## Capabilities

- Open an incident with a title, service_ids, and a severity from SEV0 to SEV4, with an Idempotency-Key header for replay-safe creates
- Acknowledge an incident, transition its status with a comment, or escalate it to the next escalation policy level
- Page a specific person by email over Slack, email, or both with a short message
- Look up the current on-call responder for a team and the escalation policy for a severity
- Add timeline comments to an incident and read its full timeline
- Write a postmortem for an incident with root_cause, resolution, action_items, and five_whys
- Read the public status summary, components, incident history, and maintenance windows without a key

## Use cases

### AI on-call assistant triaging an alert

An AI on-call assistant receiving an alert needs to open an incident and get a human looking at it. Through Jentic the agent searches for 'open an incident', loads the createIncident operation with its input schema, and calls it with the Runframe key injected at execution time. It then looks up the team's current on-call responder and pages them over Slack with a summary of the alert.

Example prompt: Call GET `/api/v1/services` with search payments to get the service id, POST `/api/v1/incidents` with title, severity SEV1, and service_ids, then GET `/api/v1/on-call/current` with team_name platform and POST `/api/v1/incidents/{id}/page` with that person's email and channels slack

### Driving an incident to resolution

During an outage the incident record should track what responders do. Runframe separates acknowledgement from status changes: POST `/api/v1/incidents/{id}/acknowledge` stamps the acknowledgement time and resolves the acknowledge SLA, while POST `/api/v1/incidents/{id}/status` handles investigate, monitor, resolve, and close with a comment of up to 500 characters. Timeline events and escalation to the next policy level have their own routes.

Example prompt: Call POST `/api/v1/incidents/{id}/acknowledge`, add a timeline comment with POST `/api/v1/incidents/{id}/events`, then POST `/api/v1/incidents/{id}/status` with the resolved status and a comment

### Postmortem drafting

After resolution the team writes up what happened. POST `/api/v1/postmortems` takes an incident_number plus fields such as summary, root_cause, resolution, impact, timeline, action_items, contributing_factors, five_whys, and executive_summary. GET `/api/v1/postmortems` returns the postmortem for an incident_number so it can be reviewed or extended.

Example prompt: Read GET `/api/v1/incidents/{id}` for the timeline, then POST `/api/v1/postmortems` with the incident_number, summary, root_cause, resolution, and three action_items

### Public status page monitoring

Customers and dependent systems check whether Runframe-hosted status pages show trouble. The public status routes return the overall summary, per-component status, incident history with a status filter and updates_limit, unresolved incidents, and maintenance windows. They need no key, so a monitor can poll them freely.

Example prompt: Call GET `/api/v1/status/incidents/unresolved` and GET `/api/v1/status/maintenance` and report any active incident or upcoming maintenance

## Key endpoints

| Method | Path | Description |
| --- | --- | --- |
| POST | `/api/v1/incidents` | Create an incident |
| GET | `/api/v1/incidents` | List incidents with filters |
| POST | `/api/v1/incidents/{id}/acknowledge` | Acknowledge an incident |
| POST | `/api/v1/incidents/{id}/status` | Transition incident status |
| POST | `/api/v1/incidents/{id}/page` | Page a person about an incident |
| GET | `/api/v1/on-call/current` | Get the current on-call responder |
| POST | `/api/v1/postmortems` | Create a postmortem |
| GET | `/api/v1/status` | Get public status summary |

## Key resources

- **Incidents** — Incidents with title, severity SEV0 to SEV4, services, assignee, timeline, and status transitions
- **On-call and escalation** — Current on-call responder per team and escalation policies per severity
- **Postmortems** — Post-incident write-ups keyed by incident_number with root cause, impact, timeline, and action items
- **Services, teams, and users** — Directory lookups used to resolve service keys, team names, and assignee emails
- **Public status** — Status summary, components, incident history, unresolved incidents, and maintenance windows

## AI readiness

This API is usable in Jentic One now. Its AI-readiness score against Jentic's framework shows where it stands today and where improvements would make it even easier for agents to use.

- **Score:** 61 / 100
- **Maturity:** AI-Aware
- **Dimensions:**
  - Foundational Compliance: 59 / 100
  - Developer Experience & Jentic Compatibility: 65 / 100
  - AI-Readiness & Agent Experience: 45 / 100
  - Agent Usability: 94 / 100
  - Security: 60 / 100
  - AI Discoverability: 72 / 100
- **View full report:** https://jentic.com/apis/runframe.io/runframe/scorecard
- **How the score is calculated:** https://docs.jentic.com/reference/api-readiness-framework/overview/
- **More about the dimensions:** https://docs.jentic.com/reference/api-readiness-framework/specification/#dimensional-model-overview

### Score it yourself

Every API in the directory is allowlisted, so you can re-score it with no key required.

- **Score your own API:** https://jentic.com/scorecard.md
- **Scoring CLI agent skill:** https://github.com/jentic/jentic-api-scorecard/blob/main/skills/jentic-api-scorecard/SKILL.md

```sh
npx @jentic/api-scorecard-cli score <openapi-url>
```

## Why Jentic

- **Setup:** Wiring Runframe by hand means attaching the rf_ bearer key, resolving service keys and user emails before you can open or assign an incident, and keeping acknowledgement, status transitions, and field edits on their separate routes. Through Jentic you install once, import Runframe from the API Directory, store the key once, and your agent calls it.
- **Permission scoping:** Runframe puts the incident id in the URL path for every action (`/api/v1/incidents/{id}/acknowledge`, /status, /page, /escalate), so a rule can bound your agent to the incident actions you pick. You choose the operations it may call, so ones that wake people up, like paging or escalation, are not included unless you add them.
- **Credential handling:** Your Runframe API key is stored once, encrypted, by your own Jentic One instance and injected at execution time. It never enters the agent's prompt, logs, or context.
- **Discovery method:** Agents search Jentic by intent such as 'open an incident' or 'who is on call', and Jentic returns the matching Runframe operation with its input schema, including the severity enum and required service_ids, so the agent calls it correctly.

## Related APIs

- **PagerDuty** — Incident response and on-call platform with alert routing
- **Rootly** — Incident management with workflows and retrospectives
- **Opsgenie** — Alerting and on-call scheduling
- **Statuspage** — Hosted status pages for customer communication

## FAQ

### What authentication does the Runframe V1 API use?

Per its OpenAPI spec, the Runframe V1 API uses HTTP bearer authentication with an API key in the rf_ format, or a session JWT, sent in the Authorization header. The incident, on-call, postmortem, and directory routes require it, while the five public status routes under `/api/v1/status` are open. GET `/api/v1/auth/verify` confirms a key and returns its scopes.

### How do I resolve an incident with the Runframe API?

Use POST `/api/v1/incidents/{id}/status` with the target status and an optional comment explaining the change. PATCH on the incident only edits fields such as title, severity, or assigned_to, and acknowledgement has its own POST `/api/v1/incidents/{id}/acknowledge` route.

### What are the rate limits for the Runframe V1 API?

The OpenAPI spec does not specify rate limits. Check the Runframe documentation at https://runframe.io for current limits; list routes such as GET `/api/v1/incidents` page at up to 100 results per request.

### Is there a Runframe MCP server?

You do not need an MCP server to give your agent Runframe. Jentic connects it directly from the API Directory: import it, store your rf_ API key once, and your agent calls the incident, paging, and postmortem operations on demand without extra tool definitions in its context.

### How do I page the on-call engineer through Jentic?

Search Jentic for 'who is on call', run GET `/api/v1/on-call/current` with the team_name, then search for 'page someone about an incident' and execute POST `/api/v1/incidents/{id}/page` with that person's email and channels such as slack. To run it on your own infrastructure, install Jentic One from its GitHub repo.

### Can I limit what my agent is allowed to do with the Runframe API?

Yes. For a triage agent, write a rule that allows creating incidents, adding timeline events, and reading on-call, so it can open and annotate incidents but cannot page people, escalate, or resolve them. Add the page route when you want it to notify responders, and every call it makes is logged.
