Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx skills add afrizzal/ratchet --skill ratchet-shipgit clone --depth 1 https://github.com/afrizzal/ratchetWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/skills/afrizzal/ratchet/ratchet-ship)<a href="https://agentmods.dev/skills/afrizzal/ratchet/ratchet-ship"><img src="https://agentmods.dev/badge/skills/afrizzal/ratchet/ratchet-ship/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/skills/afrizzal/ratchet/ratchet-ship"><img src="https://agentmods.dev/badge/skills/afrizzal/ratchet/ratchet-ship.svg" alt="Reviewed on agentmods" width="80" height="20"></a>What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00083 | $0.01312 |
| Opus 5 | $0.00042 | $0.00656 |
| Sonnet 5 | $0.00017 | $0.00262 |
| Haiku 4.5 | $0.00008 | $0.00131 |
Grade A, and why
ratchet-ship scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 12d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 55 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Ratchet Ship — land it green
Shipping is the only ratchet click users feel. Your contract: nothing leaves the machine unverified, nothing merges while CI is red, and if it goes wrong there is a written way back.
Invocation
/ratchet-ship ["summary of what's going out"]
Phase 1 — Know the pipeline before touching it
- Read the repo's release reality: CI workflows (
.github/workflows/, GitLab CI, etc.), deploy docs (README/DEPLOYMENT/CLAUDE.md), branch model (git log --graph --oneline -20, default branch, PR-based or push-to-deploy?). - If pushing the default branch triggers a production deploy, say so out loud before anything else and get explicit confirmation. Push-to-deploy repos get no casual pushes.
- Detect the local verify commands (test/typecheck/lint/build) from the manifest — and confirm each actually runs. A
lintscript that shells to a removed subcommand, or atypecheckthat only passes with a warm cache, is a broken gate wearing a green coat; verify cold. - Confirm the CI gate is enforceable, not just present. A green pipeline that cannot block a merge is advisory theater. Check branch protection / required-status-checks actually exist (
gh api repos/{owner}/{repo}/branches/{branch}/protection); if it returns 403/empty (e.g. a private free-plan repo), say so — red commits can and do reach the default branch, so your local gate is the real gate.
Phase 2 — Preflight (local, all of it)
- Run the full local gate: typecheck, tests, lint, build — whatever exists. Any red → STOP and report; shipping red is not your call to make.
- Review the outgoing diff (
git diff <base>...HEAD --stat+ read the risky hunks):- Secret scan the diff: keys, tokens, passwords, connection strings,
.envcontent. Hit → STOP immediately, tell the user (rotation may be needed if it was ever committed). - Large/binary files, lockfile churn without dependency intent, debug leftovers (
console.log, commented-out blocks, skipped tests). - Reproducibility of the CI build. If CI deletes or regenerates the lockfile and runs
npm install(notnpm ci) — or otherwise floats dependency versions — the thing CI tested is not pinned to what you reviewed. Flag it; a green run on unpinned deps is a weaker signal than it looks.
- Secret scan the diff: keys, tokens, passwords, connection strings,
- Docs sync: if the repo maintains a CHANGELOG / task ledger / ratchet BACKLOG.md, confirm it reflects what's shipping; update if the convention demands it. A v2 ratchet backlog (
<!-- ratchet:v2 -->) keeps its state inratchet/lanes/*.md, not in BACKLOG.md — read the lane files for the sync check, and never write ledger state into the (human-owned) BACKLOG.md. Gate: nopendingsha, noin-progressrow, and no openrun startedmarker (arun startedjournal line with no laterrun ended) in any lane file being shipped — an open lane is an unfinished run; STOP and name it. If lane branches are being merged as part of this ship: merge commit or fast-forward only — never squash or rebase-merge, both rewrite shas and orphan everyCommit:cell in the merged ledgers — and re-run the local gate after the merge (lanes green in isolation can be red together). - Migrations going out? Confirm they're additive or explicitly approved, and that a pre-deploy backup exists in the pipeline (or tell the user it doesn't — that's a finding, not a blocker you silently absorb).
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 12d ago First seen · 55 lines · 83 tokens per session scan A 3b227847a349
ratchet-ship is a skill published in the GitHub repository afrizzal/ratchet (6 stars, last pushed 2mo ago), licensed MIT. It adds 83 tokens to every session and 1,312 once invoked, about $0.0004 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.
Other skills, from other repositories
skill-release-management
Release planning and scheduling, semantic versioning (semver), rollout strategies (canary, blue-green, gradual), changelog generation, rollback procedures, production validation. Use when planning releases, creating rollout strategies, or managing version bumps.
github-release-management
Comprehensive GitHub release orchestration with AI swarm coordination for automated versioning, testing, deployment, and rollback management.
release
CONTRIBUTOR TOOL - Cut a plugin release: bump plugin.json version, finalize CHANGELOG, update README if needed, gate on make ci, commit, tag vX.Y.Z, and create the GitHub release. Use when shipping a new plugin version. NOT distributed.
push-ci
Push to remote and monitor CI. Validates branch safety, executes git push WITH explicit user approval, then monitors CI run status via gh CLI. Use when: user says 'push', 'push and watch CI', 'ship it', 'push-ci'. Not for: committing (use /smart-commit), creating PRs (use /create-pr), merging (use /merge-prep).
ops-ship
OPS on-demand: This skill should be used when the user asks to "ship ops plugin", "merge all PRs and…
pipeline-digest
Daily release-pipeline digest with a change gate — reports which plugins are pending release, whether main's CI is green, and which branches went stale, and notifies the operator only when that state actually moved since the last run. Use when the operator asks "what moved in the pipeline", "anything to ship"…