Borrowing it
Nothing to install: this file belongs to lglucas/ai-dev-operating-system. 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/lglucas/ai-dev-operating-system/main/.claude/skills/multi-ai-review/SKILL.mdgit clone --depth 1 https://github.com/lglucas/ai-dev-operating-systemWrote 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/lglucas/ai-dev-operating-system/multi-ai-review)<a href="https://agentmods.dev/skills/lglucas/ai-dev-operating-system/multi-ai-review"><img src="https://agentmods.dev/badge/skills/lglucas/ai-dev-operating-system/multi-ai-review/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/lglucas/ai-dev-operating-system/multi-ai-review"><img src="https://agentmods.dev/badge/skills/lglucas/ai-dev-operating-system/multi-ai-review.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 | $0.00091 | $0.01606 |
| Opus 5 | $0.00046 | $0.00803 |
| Sonnet 5 | $0.00018 | $0.00321 |
| Haiku 4.5 | $0.00009 | $0.00161 |
Grade A, and why
multi-ai-review 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 9d 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 — 147 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Multi-AI Review
Solo reasoning ships solo blind spots. For decisions with hard-to-reverse blast radius, get at least one independent reviewer with a different prior and explicitly reconcile disagreements before committing.
This is not ceremony. It is the cheapest mistake-prevention you can buy in an AI-assisted workflow: minutes of compute against hours-to-days of cleanup.
When to invoke
Mandatory:
- Architecture decisions touching auth, data model, or payments.
- Migrations, schema changes, or anything irreversible without a backup restore.
- Pricing or packaging changes (also triggers
business-plan-impact-review). - Public-facing legal/compliance text (Privacy Policy, Terms, refund policy, LGPD/GDPR statements).
- Any deploy gate where rollback is non-trivial.
- Adoption of a new external skill bundle (
#agents-marketplacepacks). - The user explicitly says "isso é importante" / "não posso errar aqui" / "double-check isso".
Optional but recommended:
- New feature spec right before sprint commitment.
- Naming a public-facing thing you can't easily rename later (product name, API endpoint, slug).
- Choosing between two non-trivially-different stacks.
Skip:
- Routine code edits, refactors, bug fixes with tests.
- Reversible UI tweaks.
- Internal-only docs.
Reviewer roles (pick at least 2)
The skill orchestrates multiple reviewers, each with a distinct lens. The user picks the relevant 2–4 per decision. Default set in bold.
| Role | Lens | Existing agent (if any) |
|---|---|---|
| Devil's advocate | "Why is this wrong?" | .claude/agents/devils-advocate-agent.md |
| Technical/security red team | "How does this break or get exploited?" | .claude/agents/technical-security-red-team-agent.md |
| Business red team | "Why does this not make business sense?" | .claude/agents/business-red-team-agent.md |
| Privacy/compliance | "Where does this leak personal data or trip LGPD/GDPR?" | uses privacy-audit skill |
| Cost watchdog | "Where does this silently get expensive?" | uses cost-watchdog skill |
| Plain-Portuguese reader | "Would a non-dev customer understand this?" | uses plain-portuguese-explainer skill |
| Independent model | Same prompt, different model family (e.g. Sonnet ↔ Opus, or external /codex) |
external |
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.
- 9d ago First seen · 147 lines · 91 tokens per session scan A 6f0e87eb2ee0
multi-ai-review is a skill published in the GitHub repository lglucas/ai-dev-operating-system (11 stars, last pushed 1mo ago), licensed MIT. It adds 91 tokens to every session and 1,606 once invoked, about $0.0005 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
cn-check
Install and run the Continue CLI (cn) to execute AI agent checks on local code changes. Use when asked to "run checks", "lint with AI", "review my changes with cn", or set up Continue CI locally.
phone-harness
Control the user's phone — iPhone through the Mac's iPhone Mirroring window, or an Android over adb: open apps, tap, type, swipe, read the screen.
security-audit
Java security checklist covering OWASP Top 10, input validation, injection prevention, and secure coding. Works with Spring, Quarkus, Jakarta EE, and plain Java. Use when reviewing code security, before releases, or when user asks about vulnerabilities.
architecture-review
Analyze Java project architecture at macro level - package structure, module boundaries, dependency direction, and layering. Use when user asks "review architecture", "check structure", "package organization", or when evaluating if a codebase follows clean architecture principles.
issue-triage
Triage and categorize GitHub issues with priority labels. Use when user says "triage issues", "check issues", "review open issues", or during regular maintenance of GitHub issue backlog.
git-commit
Generate conventional commit messages for Java projects. Use when user says "commit", "create commit", "commit changes", or after completing code changes that need to be committed.