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 mgiovani/cc-arsenal --skill team-reviewgit clone --depth 1 https://github.com/mgiovani/cc-arsenalWrote 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/mgiovani/cc-arsenal/team-review)<a href="https://agentmods.dev/skills/mgiovani/cc-arsenal/team-review"><img src="https://agentmods.dev/badge/skills/mgiovani/cc-arsenal/team-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/mgiovani/cc-arsenal/team-review"><img src="https://agentmods.dev/badge/skills/mgiovani/cc-arsenal/team-review.svg" alt="Reviewed on agentmods" width="80" height="20"></a>- NVIDIA SkillSpector pass
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 | $0.00130 | $0.03988 |
| Opus 5 | $0.00065 | $0.01994 |
| Sonnet 5 | $0.00026 | $0.00798 |
| Haiku 4.5 | $0.00013 | $0.00399 |
Grade A, and why
team-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 10d 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 — 376 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Team Review
Multi-agent team orchestration for comprehensive PR code review. Spawns 6 specialized reviewer agents plus 1 adversary reviewer as a coordinated team. Designed for security-sensitive, architectural, or high-impact code changes where a single-agent review is insufficient.
For simpler reviews, use /review-code (single-agent with parallel Explore subagents).
Prerequisites
Full mode spawns 7 named reviewer agents that message each other directly via SendMessage (the adversary needs this to cross-examine the other 6). Named agent-team spawning requires the experimental flag. Add to your environment or settings.json:
CLAUDE_CODE_EXPERIMENTAL_AGENT_TEAMS=1
Lite mode (--lite flag, or automatic fallback when the flag is unset) needs no flag: it spawns 4 combined-role Task subagents that report results back to the orchestrator instead of talking to each other directly. Fewer agents, lower cost, still covers all 7 dimensions.
No Task/subagent tools at all? Run the review yourself, sequentially, as the lead: the 7 dimensions are the methodology, parallel agents are just how Claude Code speeds it up:
- Read every file in scope (for PR/commit reviews: the diff, plus surrounding code for context).
- Work each checklist in turn, logging findings as you go: architecture → security → performance → testing → style → docs/UX (checklists are in references/agent-catalog.md, one per dimension).
- Re-read your own findings as the adversary would: which look like false positives? What's a blind spot across dimensions? What breaks at 10x scale or under adversarial input?
- Consolidate and write the report: Phases 4-5 below apply unchanged regardless of how the findings were gathered.
Delegate mode (recommended for full mode): Press Shift+Tab to enable delegate mode, which restricts the lead to coordination-only tools and prevents it from reviewing code itself.
Input
$ARGUMENTS
Notes
/resumedoes not restore teammates: an interrupted full-mode session loses the team; re-run from scratch.- Only one team-review can run per session.
- Analysis only: it finds issues but never edits code; use
/implement-featureor/fix-bugfor the actual fixes. - Teammates sometimes forget to mark tasks complete, so the orchestrator should poll task status rather than assume completion.
- Every reviewer must ground findings in code it actually read (file:line, real snippet) and flag uncertain cases as "needs manual verification" instead of asserting, a false positive here costs more than a missed finding, since it erodes trust in the whole report.
- Abort early if the requested scope (PR/commit) doesn't exist or isn't readable, don't let reviewers spin on a bad input. If reviewer tasks stall past ~10 min, or the adversary doesn't report back, proceed to consolidation with what's in hand rather than blocking the whole review.
What ships with it
4 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.
- 10d ago First seen · 376 lines · 130 tokens per session scan A 0b102a70aaa5
team-review is a skill published in the GitHub repository mgiovani/cc-arsenal (8 stars, last pushed yesterday), licensed MIT. It adds 130 tokens to every session and 3,988 once invoked, about $0.0006 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 skills, from other repositories
swarm-pr-review
Run a graph-guided, tool-augmented PR review using context packing, parallel exploration, mandatory repository-agnostic risk-family coverage with dispatch scaled to diff size and risk, independent reviewer validation, critic challenge, and metrics writeback. Use for deep pull request review with low false-positive…
include-test-files-that-assert-on-behavior-being-changed-in-decl
When delegating a task affected by this skill, include.
critic-gate
Full execution protocol for MODE: CRITIC-GATE -- plan critic review, revision loops, and hard stop before execution.
tech-debt-ci-review
Codex adapter for deep technical-debt and CI-stability audits. Use when asked to find test theater, flaky tests, missing or mis-scoped tests, brittle CI/toolchain behavior, structural debt blocking green PRs, or a remediation order for opencode-swarm.
auto-review
A code-review skill that examines changed code using the TRUST 5 criteria and automated checks. It can review files, commit ranges, or changes assigned a risk level, with optional review strategies and providers.
ax-annotation
@AX code annotation workflow skill for agent-driven tag application.