| name | write-ops-log |
| description | Publish a durable debugging or infrastructure incident record to Echo. Use when a multi-step investigation reaches a conclusion, a failure needs a postmortem, or a recovery pattern should be visible across Marin projects. |
Skill: Write Ops Log
Invoke consult-echo and search the exact error, subsystem, and incident before
writing. Edit the existing entry when the same incident is continuing. Create
one entry for a distinct incident; do not turn every retry into a new record.
Draft an Open Knowledge Format document in a temporary file:
---
type: wiki-note
title: <Failure and operational consequence>
use_when: when <future symptom should retrieve this incident>
tags:
- incident
- debugging
- <repo-or-subsystem>
- severity-<low|medium|high|critical>
- <resolved|unresolved>
---
## Impact
<Affected work, duration, and user-visible consequence.>
## Cause and resolution
<Diagnostic discriminator, established cause, recovery, and verification.>
## Evidence and prevention
<Canonical issue, PR, run, dashboard, and follow-up links.>
Include the shortest chronology needed to make the diagnosis credible. Keep
raw logs and command transcripts in their source systems.
Publish with
uv run .agents/skills/consult-echo/scripts/echo.py wiki add --file <path> or
revise with
uv run .agents/skills/consult-echo/scripts/echo.py wiki edit <id> --file <path>.
Return the canonical Echo URL and link it from the associated issue or PR. Do
not add a repository debug-log file.