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 goiltpatpat/patpat --skill patpat-shipgit clone --depth 1 https://github.com/goiltpatpat/patpatWrote 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/goiltpatpat/patpat/patpat-ship)<a href="https://agentmods.dev/skills/goiltpatpat/patpat/patpat-ship"><img src="https://agentmods.dev/badge/skills/goiltpatpat/patpat/patpat-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/goiltpatpat/patpat/patpat-ship"><img src="https://agentmods.dev/badge/skills/goiltpatpat/patpat/patpat-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 | $0.00057 | $0.00964 |
| Opus 5.5 | $0.00023 | $0.00386 |
| Sonnet 5.5 | $0.00011 | $0.00193 |
| Haiku 4.5 | $0.00006 | $0.00096 |
Grade A, and why
patpat-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 2d 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 — 37 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Patpat Ship
Before following this skill, apply the skill state and mutation gate.
When invoked directly, read the operating protocol and execution graph in full. Do not load the router.
Read preserve safety and proof over proxy.
After verify and review pass, require delivery intent: explicit commit, PR, or ship language; overnight or continuation; or explicit land or merge. Explicit /patpat activation alone is not delivery authority. Also require that no higher-priority repository rule forbids delivery. Then apply default delivery: commit the in-scope diff, use a non-force push, and open or update one ready pull request unless the user opted out or the path was read-only.
Apply authorized delivery to merge, publish, or deploy. Only explicit land or merge language authorizes merge. Overnight, going to bed, or don't stop drives a verified pull request to merge-ready and stops. Do not land a real CI failure. Retrigger an infrastructure flake at most once, only within existing delivery authority for that pull request, and never by pushing an empty commit or rerunning a job that deploys, publishes, or writes externally. A repeated failure earns no further retrigger; inspect its job logs and reclassify.
Patpat prescribes no CI system, job list, or duration. CI belongs to the project: at ship time, read its policy, configuration, and recent runs, then decide which checks prove this change and how long to wait, scaling local checks to the change's risk and to what changed since the last proof. Project instructions and the user's stated preferences are inputs to that decision. Follow what the user asks for: add or tune CI when they request it, and suggest a gap rather than adding it unasked. A failing optional check is still evidence: explain it, and it still stops a Patpat merge. Scaling chooses which checks to run; every check relied on must run at the exact head or pass the durable gate, as patpat-verify defines. The gates in the delivery playbooks stay fixed: never claim a check that did not run, and never weaken a check on your own initiative. If the user asks to relax a required check, apply the pause in preserve safety and wait for their confirmation. Relaxing a policy does not waive a real failure on this head.
For CI monitoring or PR readiness work, use PR Readiness Handoff. It is the canonical source for GitHub observation, native stack discovery, watcher budgets, stale handling, and authority boundaries.
Consume durable state when it exists. Delivery proof defines what binds a commit, which gate each delivered head uses, and how slices and whole programs differ. A failed gate stops delivery; a passed gate grants no authority.
Inspect version-control state, unrelated dirty files, and secrets before any git write. Workers never ship. Re-check the remote revision after the action.
Pause for production deploy, package publish, force-push, secret rotation, and risky auth, billing, or permission changes. Explain blast radius, rollback, and verification before continuing.
What ships with it
7 files beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
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.
- 2d ago Changed 04dac6abc505
- 8d ago Changed · +5 lines b8eddc625c1d
- 19d ago Changed 9ea9ade1c101
- 24d ago Changed 556bfd8a8c44
- 26d ago Changed · +4 lines 064d6a48a3ca
- 2mo ago Changed · -2 tokens per session ea745b80b6eb
- 2mo ago First seen · 28 lines · 59 tokens per session scan A a6df3699a34d
patpat-ship is a skill published in the GitHub repository goiltpatpat/patpat (1 stars, last pushed today), licensed MIT. It adds 57 tokens to every session and 964 once invoked, about $0.0002 per session on Opus 5.5. 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.
patpat-ship compared with skills from other repositories
| Mod | Type | Stars | Last push | Licence | Scan | Tokens | Works with |
|---|---|---|---|---|---|---|---|
| patpat-shipgoiltpatpat/patpat | Skill | 1 | today | MIT | A | 57 | Claude Code, Codex |
| git-workflow-and-versioningaddyosmani/agent-skills | Skill | 115k | 8d ago | MIT | A | 131 | Claude Code |
| git-opsPrismer-AI/PrismerCloud | Skill | 1.4k | 18d ago | MIT | A | 62 | Claude Code |
| changelog-generatorborghei/Claude-Skills | Skill | 1.1k | yesterday | none | A | 48 | Claude Code |
| git-authoringn-shadloo/git-authoring | Skill | 4 | 17d ago | MIT | A | 225 | Claude Code |
| handoffheliohq/ship | Skill | 94 | 4mo ago | MIT | A | 56 | Claude Code |
| wednesday-gitwednesday-solutions/ai-agent-skills | Skill | 173 | 5mo ago | MIT | A | 35 | Claude Code, Codex |