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 agentmods add skills/ovid/paad/releasenpx skills add Ovid/paad --skill releasegit clone --depth 1 https://github.com/Ovid/paadWrote 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/ovid/paad/release)<a href="https://agentmods.dev/skills/ovid/paad/release"><img src="https://agentmods.dev/badge/skills/ovid/paad/release.svg" alt="Measured on agentmods" 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 | $0.00034 | $0.03499 |
| Opus 5 | $0.00017 | $0.01750 |
| Sonnet 5 | $0.00007 | $0.00700 |
| Haiku 4.5 | $0.00003 | $0.00350 |
Grade A, and why
release 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 4d 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 — 300 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Start
Announce: "Preparing a paad release…"
Read @CLAUDE.md.
make release and make tag own everything mechanical: the promotion of
preview/paad into plugins/paad, the changelog roll, the version rewrite across
every manifest and announce line in both trees, the export regeneration, the full
check suite, and the annotated tag. This skill owns the parts a Makefile cannot
decide — which number to bump to, whether a release is wanted at all, and whether
the thing users install actually works afterwards.
Never hand-edit plugin.json, marketplace.json, package.json, a SKILL.md
announce line, or the changelog's version headings and link refs. Those are
generated. If the generated result is wrong, fix the generator.
Never hand-edit plugins/paad at all. It is written only by promotion. A
hand edit there does not survive: make release opens with make promote, whose
rsync -a --delete copies the pre-fix preview/paad straight over it. Nothing
catches that — the tree is committed so the dirty guard passes, and the result is
self-consistent so every check passes. The release ships and the changelog
announces a fix that is not in it. Fixes go in preview/paad, hotfixes included.
digraph release {
"Read [Unreleased]" [shape=box];
"Anything there?" [shape=diamond];
"STOP — ask whether a release is wanted" [shape=box, style=bold];
"Entries within the length rule?" [shape=diamond];
"STOP — trim first, as its own commit" [shape=box, style=bold];
"Propose MAJOR/MINOR/PATCH with reasoning" [shape=box];
"User confirms version?" [shape=diamond];
"Every preview skill in README?" [shape=diamond];
"STOP — add the README entry as its own commit" [shape=box, style=bold];
"On a branch?" [shape=diamond];
"Create release branch" [shape=box];
"make release VERSION=X.Y.Z" [shape=box];
"Passed?" [shape=diamond];
"STOP — report the failure and the half-applied state" [shape=box, style=bold];
"Read git diff plugins/ — the payload" [shape=box];
"Show the diff, get sign-off" [shape=box];
"Commit, merge to main, push" [shape=box];
"make tag" [shape=box];
"Hand off in-app verification" [shape=doublecircle];
"Read [Unreleased]" -> "Anything there?";
"Anything there?" -> "STOP — ask whether a release is wanted" [label="empty"];
"Anything there?" -> "Entries within the length rule?" [label="entries"];
"Entries within the length rule?" -> "STOP — trim first, as its own commit" [label="no"];
"Entries within the length rule?" -> "Propose MAJOR/MINOR/PATCH with reasoning" [label="yes"];
"Propose MAJOR/MINOR/PATCH with reasoning" -> "User confirms version?";
"User confirms version?" -> "Propose MAJOR/MINOR/PATCH with reasoning" [label="no"];
"User confirms version?" -> "Every preview skill in README?" [label="yes"];
"Every preview skill in README?" -> "STOP — add the README entry as its own commit" [label="no"];
"Every preview skill in README?" -> "On a branch?" [label="yes"];
"On a branch?" -> "Create release branch" [label="on main"];
"On a branch?" -> "make release VERSION=X.Y.Z" [label="on a branch"];
"Create release branch" -> "make release VERSION=X.Y.Z";
"make release VERSION=X.Y.Z" -> "Passed?";
"Passed?" -> "STOP — report the failure and the half-applied state" [label="no"];
"Passed?" -> "Read git diff plugins/ — the payload" [label="yes"];
"Read git diff plugins/ — the payload" -> "Show the diff, get sign-off";
"Show the diff, get sign-off" -> "Commit, merge to main, push";
"Commit, merge to main, push" -> "make tag";
"make tag" -> "Hand off in-app verification";
}
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.
- 4d ago First seen · 300 lines · 34 tokens per session scan A c519f75c77ed
release is a skill published in the GitHub repository Ovid/paad (109 stars, last pushed today), licensed MIT. It adds 34 tokens to every session and 3,499 once invoked, about $0.0002 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 2026-08-30.
Other skills, from other repositories
nopua
The anti-PUA. Drives AI with wisdom, trust, and inner motivation instead of fear and threats. Activates on: task failed 2+ times, about to give up, suggesting user do it manually, blaming environment unverified, stuck in loops, passive behavior, or user frustration ('try harder', 'figure it out', '换个方法', '为什么还不行').…
nopua-lite
NoPUA Lite — core wisdom in 1.5k tokens. Drives AI with trust and inner motivation instead of fear. Same Daoist philosophy, minimal footprint. For personal use and small-context models.
ccg-annotate
AI-driven annotation workflow for adding structured metadata to code. Annotations are indexed and searchable via FTS.
ccg-namespace
Isolate read-only CCG search, documentation discovery, and analysis by namespace. Use when working across multiple repositories or services, preventing cross-project graph leakage, federating supported reads, or traversing materialized cross-namespace references. Do not use for graph writes or scoped updates; the user…
ccg-docs
Generate Markdown documentation from code graphs, build RAG indexes for AI consumption, and lint documentation quality.
Codex-api
Reference for the Codex API / Anthropic SDK — model ids, pricing, params, streaming, tool use, MCP, agents, caching, token counting, model migration. TRIGGER — read BEFORE opening the target file; don't skip because it "looks like a one-liner" — whenever: the prompt names Codex/Anthropic in any form (Codex, Anthropic…