Install
$ agentstack add skill-hatch3r-hatch3r-h4tcher-pr-resolve ✓ scanned · ✓ verified, works with Claude Code, Cursor, and more.
Security review
✓ PassedNo issues found. Passed automated security review. · v0.1.0 How review works →
- ✓ Prompt-injection patterns
- ✓ Secret / credential exfiltration
- ✓ Dangerous shell & filesystem operations
- ✓ Untrusted network calls
- ✓ Known-malicious package signatures
What it can access
- ✓ Network access No
- ✓ Filesystem access No
- ✓ Shell / process execution No
- ✓ Environment & secrets No
- ✓ Dynamic code execution No
From automated source analysis of v0.1.0. “Used” means the capability is present in the source — more access means more to trust, not that it’s unsafe.
Verified badge
Passed review? Show it. Paste this badge into your README, it links to the public security report.
Reliability & compatibility
Declared compatibility
Compatibility is declared by the source manifest. End-to-end runtime verification is coming, see below.
We're building live execution health for every listing: tool-call success rate, median latency, uptime, and last-checked timestamps, measured, not self-reported. It isn't live yet, so we don't show numbers we can't stand behind.
How agent discovery & health will work →About
> Last updated: 2026-06-06
PR Resolve (Maintainer)
Maintainer-facing wrapper around commands/hatch3r-pr-resolve.md for resolving PR comments on the hatch3r repo itself. Invokes the canonical workflow through reply-posting, layers in hatch3r-development quality gates, then stops before commit so the maintainer reviews the diff.
For non-hatch3r repos the canonical /hatch3r-pr-resolve command (distributed via adapters) is the entry point — this skill is only for work on this repo.
Step 1: Preflight
- Confirm branch + open PR:
``bash git branch --show-current gh pr view --json number,title,url,state ``
- Read
governance/CONSTITUTION.md§2 P5 lean-threshold table; cache the limits for files this PR is likely to touch. - Capture baseline line counts for any governance file likely to be edited:
``bash wc -l governance/CONSTITUTION.md governance/AUDIT.md governance/AUDIT-EXECUTE.md governance/RE-ENVISION.md governance/EVOLVE.md governance/audit/domains/*.md 2>/dev/null ``
- Confirm working tree is clean:
git status --shortreturns empty.
Step 2: Invoke the Canonical Workflow
- Read
commands/hatch3r-pr-resolve.mdend-to-end. - Execute that workflow from Step 0 (Triage) through Step 8 (Post replies). Halt before Step 9 (Commit and push) — the gates below run first; the maintainer commits manually at the end of this skill.
- For every sub-agent prompt spawned during Step 6 (Fix implementation) and Step 7 (Review loop + final-quality specialists), inject this hatch3r-development context block:
``` Hatch3r-specific constraints:
- Pillar Compliance Test: every change must serve one or more of P1–P8
(governance/CONSTITUTION.md §2). If none, revert. P8 dominates P7: never under-fan-out for token-cost reasons.
- Lean thresholds: governance file line limits per CONSTITUTION §2 P5
(read the table — do not cache numeric ceilings here, they drift) — compress within the file before exceeding.
- Anti-slop wordlist: zero hits per
.claude/rules/anti-slop-enforcement.md.
- Commit format: type(scope): description with DCO sign-off via
git commit -s. `` Fan-out is task-derived (P8 B2): sub-agent count tracks the canonical workflow's Step 6/7 fan-out (implementers per fix slot, reviewer + fixer review loop, final-quality specialists). Token cost never serializes independent work (.claude/rules/fan-out-discipline.md Cost-dominance clause). Emit subagentsspawned: { count, rationale } in your output (the Sub-agents` field of the Step 5 summary).
Step 3: Hatch3r Quality Gates
Run after Step 2 fixes complete. Block commit on any failure.
| Gate | Command | Threshold | |------|---------|-----------| | Tests | npm test | 0 failed | | Typecheck | npx tsc --noEmit | 0 errors | | Lint | npm run lint | 0 errors | | Validation | npm run validate | 0 errors | | Rule parity | npm run validate:rule-parity | 0 mismatches | | Efficiency invariants | npm run validate:efficiency | 0 violations | | Build | npm run build | succeeds; dist/ produced | | Inventory drift | npm run inventory:check-docs | 0 drift |
- If canonical content was touched (
agents/,skills/,rules/,commands/,hooks/,checks/,prompts/,github-agents/):
``bash npm run inventory git diff governance/inventory.json `` Confirm the diff matches the actual content changes; investigate any unexpected drift before proceeding.
- If
src/merge/,src/content/, orsrc/adapters/customization.tswas touched, run a coverage spot-check:
``bash npm test -- --coverage ` Critical-module thresholds: 90/80/90/90 for src/merge/; 85/70/85/85 for src/content/; 85/75/85/85 for src/adapters/customization.ts`.
Step 4: Pillar + Governance Compliance
For every file modified during Step 2:
- Pillar Compliance Test (
.claude/rules/pillar-compliance.md):
- Which pillar(s) P1–P8 does this change serve?
- What measurable improvement does it produce?
- Does it increase total governance size? If yes, justify the net value or compress elsewhere.
- If no pillar served → revert the change.
- Lean threshold check: Compare new
wc -lagainst Step 1 baselines and the limits cached from CONSTITUTION.md §2 P5. Any file over limit → compress within the file.
- Anti-slop scan on every modified
.md/.mdcfile undergovernance/,agents/,commands/,rules/,skills/,hooks/:
``bash git diff --name-only HEAD | grep -E '\.(md|mdc)$' | \ xargs -I {} grep -Hn -E '(best possible|best-in-class|world-class|comprehensive and thorough|robust and resilient|it is important to note|this section describes)' {} ` Any hit → fix before commit using the replacement column from .claude/rules/anti-slop-enforcement.md`.
Step 5: Iteration Summary — STOP Before Commit
- Emit the summary below to the conversation. Do not commit, push, or merge.
PR-Resolve Iteration Summary
----------------------------
PR: # —
Branch:
URL:
Findings:
Addressed:
Deferred: (written to todo.md per canonical Step 5c)
Blocked:
Files changed:
Diff summary:
Gates:
npm test
npx tsc --noEmit
npm run lint
npm run validate
npm run validate:rule-parity
npm run validate:efficiency
npm run build
npm run inventory:check-docs
Pillars served: P, P, ...
Lean threshold deltas: after / limit>
Anti-slop hits: (must be 0 to commit)
Inventory regenerated:
Sub-agents: count=, rationale=
Suggested commit message:
():
Pillars: P, P
Status: SUCCESS | PARTIAL | BLOCKED
Confidence: high | medium | low
Next action (run manually):
git add
git commit -s -m ""
git push
- Per-comment replies were already posted in Step 8 of the canonical workflow, so the PR thread reflects the work even though the local commit is pending. The maintainer's manual commit + push completes the loop.
Constraints
- Do not edit
governance/inventory.jsonby hand — only vianpm run inventory. - Do not skip the anti-slop scan; one banned phrase blocks commit.
- Do not push to
maindirectly; commits land on the current feature or release branch. - Do not amend prior commits unless the maintainer explicitly instructs.
- Always run the Pillar Compliance Test on every governance file edit before saving.
- DCO sign-off is required:
git commit -s.
References
- Canonical workflow:
commands/hatch3r-pr-resolve.md - Pillar definitions:
governance/CONSTITUTION.md§2 (P1–P8) - Lean thresholds:
governance/CONSTITUTION.md§2 P5 - Anti-slop wordlist:
.claude/rules/anti-slop-enforcement.md - Pillar Compliance Test:
.claude/rules/pillar-compliance.md - Commit conventions + DCO:
.claude/rules/commit-conventions.md - Companion skill:
.claude/skills/h4tcher-release-prep/SKILL.md(post-merge release cuts)
Source & license
This open-source skill is cataloged on AgentStack and links to its original source — we do not rehost the code.
- Author: hatch3r
- Source: hatch3r/hatch3r
- License: MIT
- Homepage: https://docs.hatch3r.com
Install and usage instructions live in the source repository linked above.
Reviews
No reviews yet, be the first.
Write a review
Versions
- v0.1.0 Imported from the upstream source.