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 instructions/nick-railsback/skill-engine/claude-mdgit clone --depth 1 https://github.com/nick-railsback/skill-engineWrote 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/instructions/nick-railsback/skill-engine/claude-md)<a href="https://agentmods.dev/instructions/nick-railsback/skill-engine/claude-md"><img src="https://agentmods.dev/badge/instructions/nick-railsback/skill-engine/claude-md.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 | $0.00518 | $0.00518 |
| Opus 5 | $0.00259 | $0.00259 |
| Sonnet 5 | $0.00104 | $0.00104 |
| Haiku 4.5 | $0.00052 | $0.00052 |
Grade A, and why
skill-engine CLAUDE.md 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 3d 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 — 43 lines — stays where its author put it; the contents beside it link to each section on GitHub.
skill-engine — standing rules
Domain knowledge is not here. It loads on demand from
.claude/skills/skill-engine-context/ — the artifact contract, the reference
invariants, verify.sh's checks, the workflows. Keep this file to rules that
apply on every turn; anything you'd have to look up belongs in that skill.
Before proposing a commit
Run make ci-local. It is the complete local check — the same five suites
.github/workflows/lint.yml runs (shellcheck, json, doctrine, tests,
examples), with no network and no model calls. make hooks-audit is one of
those suites, not a substitute for the target.
Green here is not green in CI. ci-local covers lint.yml and
hooks-audit.yml in full. security.yml is a third workflow, and three of its
gates run nowhere locally: bandit --severity-level high, semgrep --severity ERROR, and gitleaks. They stay out on purpose — semgrep resolves p/ci and
p/secrets from the registry, so folding them in would cost ci-local the
no-network property that makes it cheap enough to run on every change. That
residual is to be absorbed, not forgotten: when a change touches Python under
plugin/skill-engine/tests/, run the two gates by hand before proposing the
commit.
bandit -r plugin/skill-engine/tests --severity-level high -q
semgrep scan --config p/ci --config p/secrets \
--config .semgrep/skill-content.yml --severity ERROR --error
Versions
Six version surfaces across five files move together, and /release is what
moves them. Never bump one by hand: doctrine.sh's version-parity check fails
the build when they disagree, which is the point.
Git
Nothing is committed, pushed, tagged, or released on the maintainer's behalf —
prepare the change, then surface the commands for them to run. git commit,
git push, git tag and gh release create are all ask-gated in
.claude/settings.json, so the rule has mechanical backing; keep the two in
step if either changes. git add is not ask-gated — staging isn't
publishing, and re-confirming it on every save was pure friction.
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.
- 3d ago First seen · 43 lines · 518 tokens per session scan A 082712aa277d
skill-engine CLAUDE.md is an instructions file published in the GitHub repository nick-railsback/skill-engine (2 stars, last pushed 26d ago), licensed MIT. It adds 518 tokens to every session, about $0.0026 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 instructions, from other repositories
erpaval CLAUDE.md
Instructions for theagenticguy/erpaval, covering erpaval-plugin, repo structure, version management, tools and hooks.
Circuit AGENTS.md
Instructions for CIrcui-try/Circuit, covering 슬래시 커맨드 공통 규칙, claude / codex 양립, plan mode 호환 and 커밋 / github 계정.
Bamboo CLAUDE.md
Claude Code instructions for VolcanoEngineering/Bamboo, covering claude.md, what this repo is, cold-start order for claude, claude-shaped operating notes and when you're working on this repo.
Copilot-Skills git-ops.instructions.md
Auto-loaded context for Git version control operations: commits, merges, rebases, and GitHub workflows.
Copilot-Skills lazycommit.instructions.md
Auto-loaded context for lazycommit skill: AI-powered commit message generation with diff analysis.
vscode buildNext.instructions.md
Working notes and architecture documentation for the new esbuild-based build system in build/next. Use when making changes to the new build pipeline (transpile/bundle commands, NLS plugin, source-map handling, resource copying, or self-hosting watch tasks).