Jaime

Learn about actions >

• ask

  Append one conversational question to the current (or a given) incident and return a new AI answer as a turn. Never overwrites previous answers.

  Params

  • incident-id string

    Incident to ask about. Empty uses the single open incident; required when several are open, and to ask about a closed incident.

  • question string

    The question to ask about the incident (required).

• collect-context

  Write a bounded context bundle to disk and return its path.

• generate-report

  Collect fresh context and write a new report version for the current open incident (<incident-id>-<UTC timestamp>.md). Earlier versions are never overwritten; the new one becomes the report ask and get-suggestion use. Fails if no incident is open.

• get-suggestion

  Return the AI suggestion for the current open incident. Fails if no incident is open or no suggestion has been generated yet.

  Params

  • additional-context string

    Extra context to include in the AI prompt (e.g. juju config, replication status).

  • force boolean

    Call the AI provider again even when a suggestion for this context and model is already stored.

• list-incidents

  List incidents recorded in the audit log, newest first. Each entry correlates incident-start, report-generated and incident-closed events by incident id, including open incidents. Optionally filter by unit.

  Params

  • unit string

    Restrict the result to incidents for one unit (e.g. postgresql/0). Empty returns all incidents.

• list-suggestions

  Return the full conversation (all turns) for an incident as a JSON array under 'result'.

  Params

  • incident-id string

    Incident ID to list turns for. Empty lists every tracked incident's turns.

• reset

  Clear all tracked incidents and status state. Use to start fresh after a resolved incident.

• show-status

  Show the current monitoring state for every monitored unit: workload, active incident if any, and last observation. Returns a JSON array under 'result'.

• show-usage

  Show AI token usage and cost. Returns a global summary across all incidents, or per-incident detail if incident-id is provided.

  Params

  • incident-id string

    Incident ID to filter by. If empty, returns the global summary.