hermes - 💡(How to fix) Fix Skill proposal: auditable decision trail for long autonomous coding runs

Official PRs (…)
ON THIS PAGE

Recommended Tools

×6

Utilities matched from this issue’s tags and category — try them while you read without losing context.

GitHub issue graph ai analysis

Paste a GitHub issue URL. We fetch that issue, discover linked issues from bodies/comments/timeline, collect linked pull requests, and produce a structured English report.

The report is written in English Markdown for sharing and archival.

Helpful · Quick feedback

Loading…
RAW_BUFFERClick to expand / collapse

Source pattern: cursor/plugins pstack at 21327be, especially show-me-your-work and figure-it-out.

Hermes adaptation:

  • Add a lightweight optional decision-trail artifact for long-running coding/kanban/cron work.
  • Default local path could be .hermes/audit/<task-or-run>.tsv or a Kanban-run attachment, with columns: timestamp, phase, decision, why, evidence, result.
  • Append-only by convention; evidence must point to a commit, PR, file:line, test log, screenshot, CI run, or generated artifact.
  • Before final handoff, audit rows against actual tool outputs/session events and remove/update unsupported claims.
  • Commit or attach the trail only when review trust requires it; keep it local for routine work.

Why this is useful:

  • Hermes already runs unattended cron and Kanban workers. The user sees final summaries, but not always the decision trail behind pivots, rejected attempts, or partial verification.
  • A small structured artifact improves reviewability without installing Cursor plugin runtime files.
  • This complements, rather than replaces, issue #25380 worker-session mapping and issue #26742 claim verification.

Acceptance criteria:

  • A bundled skill or coding-ops doc defines the TSV schema and when to use it.
  • Kanban/cron/coding workflows can name the trail path in final summaries.
  • The guidance requires evidence pointers and forbids aspirational or invented rows.
  • At least one test or fixture validates log-row sanitation for tabs/newlines and spreadsheet formula prefixes if a helper script is added.

Vote matrix · Quick signals

Works
Did the solution work? Tap to confirm.
Easy Fix
Was it a quick fix?
Time Saver
Did it save you time?
Blocking
Was it severely blocking?
Common Issue
Are others likely hitting this too?
Flaky / Intermittent
Is it intermittent?
Verified / Reproducible
Can you reproduce it reliably?
Loading…

Still need to ship something?

×6

Another batch ranked right after the header list — different links, same matching logic.

Back to top recommendations

TRENDING