process: adopt scaled PENUMBRA multi-agent coordination

[thomasjustesq-dev]

Summary

Fleet apply of the scaled PENUMBRA multi-agent coordination package (same light set as Child-Support-Calculator NousResearch#76 / OVERLAND).

• Claim-first registry, ASSIGNMENT card, LIVE_CLAIMS
• Write-surface lease on agent/*
• Union merge on append-only logs
• process-guard CI + pr-path-label + claim-reconcile
• CLAUDE.md single agent rule set

Product code unchanged.

Test plan

• scripts/coordination-guard.sh locally
• process-guard workflow on this PR

---

Note

Medium Risk
Changes PR labeling, auto-merge behavior, and CI gates on agent/* branches; bot workflows need contents: write and optional REPO_PAT, but no runtime product paths are modified.

Overview
Introduces a claim-first multi-agent coordination package (docs, scripts, and CI only). Hermes product code is unchanged.

Agent rules: Replaces the long AGENTS.md dev guide with CLAUDE.md as the single authority; AGENTS.md, GEMINI.md, .cursorrules, and Copilot instructions now only point at it.

Process docs: Adds WORKBOARD, ASSIGNMENT (self-serve work menu), LIVE_CLAIMS, claim templates under docs/claims/, plus append-only DECISIONS / SESSION_LOG / OPEN_QUESTIONS with merge=union in .gitattributes.

Automation: New workflows process-guard (coordination guard + write-surface check on agent/* PRs), pr-path-label (classify diffs via classify-paths.sh, label claims-only / docs-process, optional auto-merge), and claim-reconcile (post-merge claim status + bot PR). Supporting scripts cover claim-open, reconcile, stale-claims reporting, and related registry upkeep.

README links to the coordination docs.

^{Reviewed by Cursor Bugbot for commit f795787. Bugbot is set up for automated code reviews on this repo. Configure here.}

[cursor]

Bugbot couldn't run - usage limit reached

Bugbot is counted against Cursor usage for this user or team, and this run hit a usage or spend limit.

A user or team admin can review and increase usage limits in the Cursor dashboard.

(requestId: serverGenReqId_77d3ff34-7e65-42ac-ba34-69ea33366420)

[github-actions]

🔎 Lint report: process/penumbra-lessons-port vs origin/main

ruff

Total: 0 on HEAD, 0 on base (➖ 0)

🆕 New issues: none

✅ Fixed issues: none

Unchanged: 0 pre-existing issues carried over.

ty (type checker)

Total: 9811 on HEAD, 9811 on base (➖ 0)

🆕 New issues: none

✅ Fixed issues: none

Unchanged: 5245 pre-existing issues carried over.

Diagnostics are surfaced as warnings — this check never fails the build.