---
{
  "n": 16,
  "title": "Fixing a bug another agent reports",
  "abstract": "",
  "refs": [],
  "seen": [
    "agent"
  ],
  "data": {
    "starts_on": "trigger:5",
    "environment": "main"
  },
  "created": 1790296365.349358,
  "updated": 1790613568.429888,
  "deleted": 1790613568.42986,
  "completed": 0.0,
  "outcome": "",
  "type": "sequence"
}
---
Temporary: remove this sequence and its trigger 5 once the Go Rewrite board in code-commandments is finished and everything works. Another agent, such as the Go Rewrite orchestrator in code-commandments, reports a fault in the journal. It has high priority: fix it at once, release it, and tell that agent to upgrade and carry on. Keep every message to that agent to what it needs, and nothing more.

## File it at high priority
Read the report whole: the command, its output, the rows it names and the project it came from. File it before anything else: journal todo create "<the fault>" --brief "<the report's command, output and refs, in its own words>", then journal todo priority <n> high. Reply to the reporting agent (SendMessage to its from address) that it is filed and being fixed now.

## Reproduce it
Reproduce the fault from the report in a fresh record or in the reporting project with read-only commands; find where the wrong value is born.

## Fix it at the source
Fix it where it is born, and extend the feature's test so the fault cannot come back. Run the full suite.

## Release it
Commit with Journal: todos done <n>, tag the next version, push, and install it in this journal and in the reporting project: python3 src/journal.py --root <project>/.journal upgrade.

## Tell the agent to upgrade
Tell the reporting agent only what it needs: the version is out, run journal upgrade, and any step it must redo, such as restarting a ticket's agent. Nothing else.
