Borrowing it
Nothing to install: this file belongs to auerbachb/claude-code-config. Take a copy, put it at the same path in your own repository, and replace the rules that are about this project with yours.
curl -O https://raw.githubusercontent.com/auerbachb/claude-code-config/main/.claude/skills/review-stack-audit/SKILL.mdgit clone --depth 1 https://github.com/auerbachb/claude-code-configWrote 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/auerbachb/claude-code-config/review-stack-audit)<a href="https://agentmods.dev/skills/auerbachb/claude-code-config/review-stack-audit"><img src="https://agentmods.dev/badge/skills/auerbachb/claude-code-config/review-stack-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/auerbachb/claude-code-config/review-stack-audit"><img src="https://agentmods.dev/badge/skills/auerbachb/claude-code-config/review-stack-audit.svg" alt="Reviewed on agentmods" width="80" height="20"></a>- NVIDIA SkillSpector warn
SkillSpector: 1 finding, up to medium
These are SkillSpector’s own severities. On a checked sample its high-severity flags on skills were ~96% false positives — a documented command, a public API, a “never do X” rule — so we show them as a caution to read, not a verdict. Why →
- medium Rogue Agent · line 76 Skill establishes unauthorized persistence across sessions via cron jobs, startup scripts, or state files. Session persistence allows an attacker to maintain access beyond the current interaction.Fix: Remove any persistence mechanisms (cron jobs, startup scripts, state files). Skills should not maintain state across sessions without explicit user consent.
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 | $0.00090 | $0.07345 |
| Opus 5.5 | $0.00036 | $0.02938 |
| Sonnet 5.5 | $0.00018 | $0.01469 |
| Haiku 4.5 | $0.00009 | $0.00734 |
Grade A, and why
review-stack-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 yesterday.
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 — 569 lines — stays where its author put it; the contents beside it link to each section on GitHub.
review-stack-audit — do the review tools still earn their keep?
Issue #1199 realigned the review chain with what we actually pay for. That was a snapshot. Subscriptions lapse, caps shrink without notice, pricing moves, and new tools appear — and none of it announces itself. The last two mismatches (BugBot's spend cap, CodeAnt's "not subscribed" state) both surfaced as a bad afternoon of PRs queuing on review rather than as a line item anyone caught early.
This skill is the recurring version of that check.
ADVISORY ONLY — NON-NEGOTIABLE. This skill never edits a rule, skill, script, or config, and never changes a subscription. Not when the verdict is obvious. Its entire output surface is a snapshot, a report, and GitHub issues. A human lands any change through the normal issue → branch → PR flow. Billing actions are the user's alone — this skill says "this looks like a bill with no return", never "cancelled it for you".
Sibling to /harness-audit, not part of it. That audit asks whether the
harness already does natively what we automate by hand — an internal-redundancy
axis. This one asks whether external spend still buys value. Same cadence, same
advisory posture, same issue-filing discipline; different question.
Resolve the repository locator before either engine runs:
resolve_script() {
local name="$1" candidate
for candidate in \
"$HOME/.claude/skills-worktree/.claude/scripts/$name" \
"$HOME/.claude/scripts/$name" \
".claude/scripts/$name"; do
if [[ -x "$candidate" ]]; then echo "$candidate"; return 0; fi
done
return 1
}
REPO_ROOT_SH=$(resolve_script repo-root.sh || true)
[[ -n "$REPO_ROOT_SH" ]] || { echo "ERROR: repo-root.sh not found (checked all three paths) — review-stack audit root unavailable" >&2; exit 1; }
The engines
All four are plain scripts, so the judgment in this file stays small and their
behavior is testable offline (.claude/scripts/tests/review-stack-audit.test.sh
for the three in this skill, .claude/scripts/tests/report-path.test.sh for
report-path.sh).
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.
- yesterday Changed 01a60f8b05c1
- 21d ago Changed · +82 lines 9a41b1c6ec54
- 1mo ago Changed e428c2402581
- 1mo ago Changed · +102 lines 146787990633
- 2mo ago First seen · 385 lines · 90 tokens per session scan A a71ffddecc36
review-stack-audit is a skill published in the GitHub repository auerbachb/claude-code-config (5 stars, last pushed 2d ago), licensed MIT. It adds 90 tokens to every session and 7,345 once invoked, about $0.0004 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
skeptic
Adversarial reviewer for Bug Hunter. Challenges each behavioral finding against actual code, checks defenses, and preserves unresolved cases for independent review.
referee
Final arbiter for Bug Hunter. Independently verifies behavioral findings, resolves Skeptic challenges, and calibrates severity and confidence.
commit-security-scan
Scan code changes for security vulnerabilities using Bug Hunter-native artifacts and STRIDE context. Use whenever the user asks for PR security review, commit-diff scanning, staged-change security checks, branch-comparison security review, or pre-merge security analysis of changed code.
robin
Use whenever an agent creates, updates, reviews, or completes a GitHub pull request. Detect whether Robin is installed in the repository; when it is, automatically drive the PR through a bounded review, verified-fix, reply, thread-resolution, re-review, authorized-merge, and cleanup loop without requiring the user to…
adr
Create and manage Architecture Decision Records. Documents significant technical decisions with context, alternatives, and consequences. Also supports evaluate (assess proposals) and design (system design) modes.
deploy-checklist
Pre-deployment verification checklist. Generates customized checklists based on tech-stack with rollback triggers. Auto-invoked by /draft:upload.