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 nick-railsback/skill-engine --skill self-auditgit 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/skills/nick-railsback/skill-engine/self-audit)<a href="https://agentmods.dev/skills/nick-railsback/skill-engine/self-audit"><img src="https://agentmods.dev/badge/skills/nick-railsback/skill-engine/self-audit/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/nick-railsback/skill-engine/self-audit"><img src="https://agentmods.dev/badge/skills/nick-railsback/skill-engine/self-audit.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 | $0.00045 | $0.01584 |
| Opus 5.5 | $0.00018 | $0.00634 |
| Sonnet 5.5 | $0.00009 | $0.00317 |
| Haiku 4.5 | $0.00005 | $0.00158 |
Grade A, and why
self-audit 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 1mo 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 — 142 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Self-audit
Run the eight drift checks against the contextualizer's current state: stale date, broken URL, long-untouched reference, catalog-vs-content disagreement, cross-reference-vs-content disagreement, review-state staleness, paragraph→permalink density, grounded-citation rate. Read-only by default; after the findings table prints, offers an opt-in propose → validate → approve gate for the three deterministic checks (stale dates, catalog-row drift, review-state staleness). The other five checks remain advisory — they print a one-line recommendation each and the human acts manually.
Contextualizer root
Engine workflows operate inside a contextualizer installed as a project skill at one of three install levels:
- User-level:
~/.claude/skills/<slug>-context/ - Local-user-level:
~/.claude/local/skills/<slug>-context/(when in use) - Project-level:
<repo>/**/.claude/skills/<slug>-context/
Every path below — research/..., references/..., verify.sh —
resolves relative to whichever directory matches. Before reading
anything, locate the root by searching all three install levels in
order:
Run the script in shared/locator-block.md verbatim before proceeding.
Selecting a contextualizer
/skill-engine:self-audit <name> names the contextualizer to audit:
<name> is the directory name without the -context suffix, the same
grammar review/apply/discard use. Substitute it (or the empty
string) for <name> in the locator above. With no argument,
auto-detection applies — it succeeds when exactly one contextualizer is
installed and lists the matches and exits when more than one is.
Read every subsequent research/foo path as $CTX_ROOT/research/foo,
every references/foo as $CTX_ROOT/references/foo, and verify.sh as
$CTX_ROOT/verify.sh.
Fleet (--all)
/skill-engine:self-audit --all audits every contextualizer the locator
enumerates rather than one. Run the script in
shared/locator-block.md with --all —
and with its name="<name>" line substituted to name="", since a fleet
run names no contextualizer and the block's find -name "<name>-context"
would otherwise match nothing — to get the roots, one absolute path per
line, then run the eight drift checks
against each enumerated contextualizer in turn, in sequence — a sweep is
the single-contextualizer audit repeated, not a different audit, and the
checks stay read-only throughout.
What ships with it
3 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.
- 1mo ago Changed · +21 lines 8fa8a5c42771
- 1mo ago Changed 75960ba21855
- 2mo ago First seen · 121 lines · 45 tokens per session scan A ce2d6a46fed6
self-audit is a skill published in the GitHub repository nick-railsback/skill-engine (2 stars, last pushed 21d ago), licensed MIT. It adds 45 tokens to every session and 1,584 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.
Other skills, from other repositories
issue-analyze
GitHub Issue and PR review thread deep analysis with Codex blind verdict. Use when: analyzing issue root cause, classifying problems, investigation planning, triaging PR review comments for actionability. Not for: fixing bugs (use bug-fix), code exploration (use code-explore). Output: classified analysis + verdict…
code-explore
Pure Claude code investigation. Use when: tracing execution paths, understanding architecture, diagnosing issues. Not for: dual-perspective review (use code-investigate), code review (use codex-code-review). Output: analysis report with findings.
debug
Interactive debugging workflow with hypothesis-driven probe loop. Use when: unknown bugs, script errors, silent failures, troubleshooting. Not for: known bugs (use bug-fix), GitHub issue analysis (use issue-analyze), code understanding (use code-explore). Output: debug report with probe journal + root cause + fix.
code-investigate
Dual-perspective code investigation. Use when: deep code analysis needing both Claude and Codex perspectives. Not for: quick exploration (use code-explore), code review (use codex-code-review). Output: integrated findings from dual analysis.
git-investigate
Git history investigation. Use when: tracking code changes, finding where bugs were introduced, root cause analysis. Not for: code exploration (use code-explore), issue analysis (use issue-analyze). Output: history trace + root cause report.
simplify
Wrap-up refactoring — simplify code, eliminate duplication, preserve behavior.