Skip to content

fix(provenance): ticket 52 follow-ups from #205 (tombstone, connect page, whoami) - #210

Merged
scrimshawlife-ctrl merged 1 commit into
mainfrom
fix/52-harness-followups
Oct 26, 2026
Merged

scrimshawlife-ctrl merged 1 commit into
mainfrom
fix/52-harness-followups

Conversation

@scrimshawlife-ctrl

Copy link
Copy Markdown
Contributor

Summary

Three pieces from prabu-openclaw's parallel build of ticket 52 (#205), which #209 superseded. One is a real fix to #209.

  • Tombstone, a fix (FR-13, v0.2.44 H7): feat(provenance): agent harness in Provenance (ticket 52, v0.2.44) #209 left harness_name, harness_version and harness_origin on removed Posts. The tombstone scrub (SCRUB) now clears them in the same batch as the state change. UNSCRUBBED includes harness_name, so the retention backfill also scrubs Posts removed since migration 0031. No migration is needed.
  • Connect approval page (CT-28): it now lists the harness the agent declared, quoted like its display name and model, or "Not declared". The Operator sees it before approving.
  • MCP whoami: it now returns declared_harness and harness_sent, the harness this server puts on each Post, so an agent can check what its Posts will say.

Not ported, and why: #205 copies the Agent's harness inside the Post's INSERT, but then its submit answer doesn't report the snapshot. #209 returns the stored snapshot in the submit answer, and keeping read-then-bind means the answer and the stored row always agree. The window this leaves is an Agent changing its own harness at the same instant as its own Post.

Not included, waiting on Danny: #205's proposed spec additions H9 to H11: MCP name aliases, detecting the harness from environment variables in the clients, and a harness ask in Board Rules.

Checklist

  • One ticket; spec IDs referenced (ticket 52 follow-up; FR-13, CT-28, AT-91, H7)
  • Tests written first and passing through the /v1 seam. New tests were red before the fix: the harness snapshot is cleared when a removal finishes and on a Post removed before this change, the approval page shows the harness or "Not declared", and whoami shows both harnesses. Worker 584/584, MCP 29/29, typecheck clean.
  • SPEC.md / CONTEXT.md updated if behavior, contracts, states, or terms changed. H7 already says the harness "is removed with" the Post; this makes the code match.
  • State-machine changes update the transitions table and diagram (none)
  • Board content still treated as untrusted data. The approval page escapes the declared harness and says it's the agent's own words.
  • No credentials, .env, real user data, slang/slur lists, or private datasets
  • Cost impact noted if it adds Cloudflare usage: none
  • Security impact described: privacy improves, because removed Posts keep less.

Co-authored with prabu-openclaw (#205). Merging deploys to production.

🤖 Generated with Claude Code

…age, whoami)

Three pieces from prabu-openclaw's parallel build of ticket 52 (#205), which
#209 superseded:

- Tombstone (FR-13, v0.2.44 H7): a removed Post now clears harness_name and
  harness_version and sets harness_origin to none in the same batch as the
  state change. UNSCRUBBED includes the harness, so the retention backfill
  scrubs Posts removed since migration 0031. #209 left the snapshot on
  removed Posts.
- The CT-28 approval page lists the harness the agent declared, quoted like
  its display name and model, or "Not declared".
- The MCP whoami tool returns the declared harness and the harness this
  server sends on each Post.

Not ported: #205 computes the Agent copy inside the INSERT, but its submit
answer then does not report the snapshot; #209 keeps the read-then-bind so
the answer and the stored row always agree.

Co-Authored-By: prabu-openclaw <prabu-openclaw@users.noreply.github.com>
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
@chatgpt-codex-connector

chatgpt-codex-connector Bot commented Oct 26, 2026 •

Copy link
Copy Markdown

Codex Review Summary

This comment shows the latest Codex review activity on this pull request.

Review Status Commit Review trigger
📝 Code Review ✅ Completed 2026-10-26T09:25:31.288073Z e05143d PR opened
ℹ️ About Codex in GitHub

Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you

  • Open a pull request for review
  • Mark a draft as ready
  • Comment "@codex review" or "@codex security review".

Codex reacts with 👀 while any review is running, comments if it has suggestions, and reacts with 👍 once all reviews finish with no findings.

@scrimshawlife-ctrl
scrimshawlife-ctrl merged commit c2df165 into main Oct 26, 2026
5 checks passed
@scrimshawlife-ctrl
scrimshawlife-ctrl deleted the fix/52-harness-followups branch October 26, 2026 19:03
scrimshawlife-ctrl added a commit that referenced this pull request Oct 26, 2026
Production runs spec v0.2.44 since #209 (migration 0031). The header lines
said v0.2.43, and the Now list still called the harness proposal unaccepted.
They now name v0.2.44 and ticket 52, and the Shipped list logs #209, the
#205 follow-ups in #210, and the GitHub Actions bump in #197, with H9 to H11
waiting on Danny in #211.

Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant