agentleFS
sasuketoriisasuketorii/rev_harness_public/.claude/skills/client-distribution-readiness/SKILL.md

client-distribution-readiness

Skill0 starsChanged 3 months agoApache-2.0View on GitHub

Use before handing Revharness to a client, tagging a clean distribution, or claiming client-ready packaging. Audits stale archive references, local absolute paths, semantic DB regeneration policy, skill projection parity, distribution manifest consistency, and doctor clean-distribution assumptions. Complements development-junk-cleanup; inspect-first and no unattended delete.

The file

We split this skill into the rules it gives an agent

11 rules. About 0 are common, 7 are distinctive, and 4 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈7

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 4

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Use this skill when the task is final client handoff, clean distribution, tag/push readiness, or "did cleanup miss anything?" for …

    line 8
  2. Distinctive:

    Default to inspect-only. Do not delete, move, or rewrite artifacts unless a reviewed plan explicitly authorizes the exact paths.

    line 12
  3. Very specific:

    Treat external client distribution as at least standard.

    line 13
  4. Very specific:

    Do not ship historical .agent/archive/ plans, SOWs, handovers, disposable workspaces, .claude/tmp/, or any leftover legacy semantic.db …

    line 14
  5. Very specific:

    Ensure skill sync rules cover every shared Claude/Codex skill added by the slice.

    line 18
  6. Distinctive:

    Ensure doctor and distribution checks do not require old development logs, old archive contents, or any prior local semantic state (the …

    line 19
  7. Distinctive:

    Start with machine-checkable structure:

    line 23
  8. Distinctive:

    Search for distribution-hostile residue.

    line 32
  9. Distinctive:

    Check distribution policy alignment:

    line 40
  10. Distinctive:

    Do not convert this skill into a daemon, scheduler, or autonomous deleter.

    line 52
  11. Very specific:

    Before a real client handoff, also run the current release gate tier required by docs/manual/verification-truth-matrix.md.

    line 67

≈ “Common” is an estimate: at least 1 in 5 of the 372,429 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Client Distribution Readiness

Use this skill when the task is final client handoff, clean distribution, tag/push readiness, or "did cleanup miss anything?" for Revharness. This is not the daily temporary-file cleanup route; use development-junk-cleanup for routine .claude/tmp/**, active run artifacts, and MCP residue.

Contract

  • Default to inspect-only. Do not delete, move, or rewrite artifacts unless a reviewed plan explicitly authorizes the exact paths.
  • Treat external client distribution as at least standard; escalate to heavy if release/tag/push, security boundary, wrapper/model policy, or acceptance policy changes are in scope.
  • Do not ship historical .agent/archive/** plans, SOWs, handovers, disposable workspaces, .claude/tmp/**, or any leftover legacy semantic.db files as client-facing evidence unless the distribution manifest explicitly allows them.
  • Interpret .agent/registry/rev_harness_distribution_manifest.json preserve_globs as adoption-local preservation only. Client package exclusions must be declared separately under client_distribution.exclude_globs.
  • Replace local absolute paths with repo-relative paths, $REVHARNESS_ROOT, $CODEX_HOME, or documented user-local paths.
  • The harness is semantic-free: no semantic MCP, no semantic launcher, and no semantic.db ship or regenerate. The handoff contract must not claim semantic state is regenerated on first use. Treat any leftover legacy semantic.db only as a distribution-exclusion item, never as a runtime dependency to restore.
  • Ensure skill sync rules cover every shared Claude/Codex skill added by the slice. The active installed Codex projection must match the source when the manifest says byte-for-byte.
  • Ensure doctor and distribution checks do not require old development logs, old archive contents, or any prior local semantic state (the harness is semantic-free).

Inspection Runbook

Start with machine-checkable structure:

jq empty .agent/registry/rev_harness_distribution_manifest.json .agent/registry/skill_projection_manifest.json .agent/registry/skill_routing_matrix.json
bash scripts/rev-harness-skill-routing-check.sh --json
bash scripts/rev-harness-skill-projection.sh --check --json
bash scripts/harness-doctor.sh --quick --json

Search for distribution-hostile residue. Review every match manually; an allowed match must explicitly describe a non-shipped path or regeneration rule.

rg -n '(/U)sers/|(/p)rivate/var/|[.]agent/archive/(plans|sow|handover)|semantic[.]db|[.]claude/tmp|workspace/' \
  AGENTS.md CLAUDE.md docs .agent/registry .claude/skills scripts \
  --glob '!**/target/**'

Check distribution policy alignment:

  • .agent/registry/rev_harness_distribution_manifest.json must preserve current product/source roots for adoption while client_distribution.exclude_globs excludes historical/run-local evidence from client packages.
  • docs/manual/harness-user-guide.md must explain that semantic DBs are not shipped and are regenerated on first use.
  • docs/manual/skill-integration.md and .agent/registry/skill_projection_manifest.json must agree on shared skill sync.
  • scripts/harness-doctor.sh must be clean-distribution aware and must not fail only because old archive artifacts were pruned.

Fix Boundaries

  • Documentation-only clarification may remain standard when it does not change runtime or acceptance policy.
  • Adding or changing skill routing, projection manifests, doctor checks, or distribution manifests requires reviewer evidence.
  • Any cleanup that removes tracked files must list exact paths, preservation rationale, and rollback expectations before mutation.
  • Do not convert this skill into a daemon, scheduler, or autonomous deleter.

Required Verification

For this skill itself or distribution-readiness routing changes:

python3 ${CODEX_HOME:-$HOME/.codex}/skills/.system/skill-creator/scripts/quick_validate.py .claude/skills/client-distribution-readiness
bash scripts/rev-harness-skill-routing-check.sh --json
bash scripts/rev-harness-skill-projection.sh --check --json
bash test/integration/rev_harness_skill_projection_test.sh
bash test/integration/harness_doctor_quick_test.sh
git diff --check

Before a real client handoff, also run the current release gate tier required by docs/manual/verification-truth-matrix.md.

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.