ddadminandClaude Opus 4.6 d9c78861e2 fix(skill): устойчивость multi-round ревью при нестабильном Agent Teams
SendMessage для тиммейтов Agent Teams работает нестабильно: может быть
недоступен как инструмент или создавать нового агента без контекста.

Step 5 теперь проверяет доступность тиммейта перед отправкой. Если
тиммейт неактивен — спрашивает оператора (interactive) или завершает
со статусом "fixes applied, not re-verified" (headless). Re-spawn
доступен как опция, не как дефолт.

Также: multi-round секция в adversarial-reviewer.md, исправленные
пути установки в README, убраны ложные утверждения о statefulness.

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
2026-04-16 12:18:52 +03:00

Claude Team Review

Adversarial code and plan review using Claude Code Agent Teams.

One teammate reviews. The lead fixes. Iterate until approved.

What is this

A Claude Code skill that spawns an adversarial reviewer as an Agent Teams teammate. The reviewer reads your project, runs tests, checks documentation, and delivers findings with a skeptical stance. The lead (your main session) fixes issues and requests re-review from the same teammate. If the teammate is no longer active, the lead decides how to proceed — re-spawn or conclude.

How it differs from adversarial-review

adversarial-review uses two different models (Claude writes, Codex reviews) — you get cross-model blind spot coverage and cheap re-review via codex exec resume. It requires Codex CLI and an OpenAI API key.

claude-team-review stays within the Claude ecosystem. No external dependencies. The reviewer is a Claude Code teammate with its own context window, MCP access, and the ability to run commands. For re-review, the lead tries to continue the same teammate; if the teammate is no longer active, the lead can re-spawn or conclude based on context. The trade-off: same model family means no cross-model diversity.

Use adversarial-review when you want maximum review quality through model diversity. Use claude-team-review when you want zero external dependencies and a richer reviewer (tests, docs, web search).

How it works

┌──────────┐     spawn        ┌────────────┐
│   Lead   │ ───────────────> │  Reviewer   │
│  (code)  │                  │ (teammate)  │
└──────────┘                  └────────────┘
     ^                              │
     │          findings            │
     │ <────────────────────────────┘
     │
     │  fix issues
     v
┌──────────┐     message      ┌────────────┐
│   Lead   │ ───────────────> │  Reviewer   │
│  (fixed) │  "re-check this" │ (same / new)│
└──────────┘                  └────────────┘
                                    │
                              VERDICT: APPROVED

The lead tries to continue the same teammate for re-review. If the teammate is no longer active (Agent Teams limitation), the lead can re-spawn with a full briefing or conclude without re-verification.

Three modes

Mode What it reviews When to use
plan Implementation plan Before writing code
code Git diff (unstaged, staged, branch) After writing code
code-vs-plan Code changes against the plan Verify implementation

Mode is auto-detected from context, or you can force it with an argument.

What the reviewer can do

  • Read any file in the repository
  • Run commands — tests, linters, type checkers, build scripts
  • Search the web and query documentation via MCP (Context7)
  • Inspect git history — blame, log, diff

The reviewer cannot create, edit, or delete project files.

Requirements

  • Claude Code ≥ 2.1.32
  • Agent Teams enabled (experimental)

No external dependencies. No API keys beyond your Claude subscription.

Installation

# Clone the repository
git clone https://github.com/dementev-dev/claude-team-review.git
cd claude-team-review

# Symlink the skill
ln -s "$(pwd)" ~/.agents/skills/claude-team-review

# Symlink the reviewer agent definition
mkdir -p ~/.claude/agents
ln -s "$(pwd)/adversarial-reviewer.md" ~/.claude/agents/adversarial-reviewer.md

Enable Agent Teams in your Claude Code settings:

{
  "env": {
    "CLAUDE_CODE_EXPERIMENTAL_AGENT_TEAMS": "1"
  }
}

Restart Claude Code after installation for the skill to be recognized.

Usage

# Auto-detect what to review
/claude-team-review

# Review a plan
/claude-team-review plan

# Review code changes
/claude-team-review code

# Review a specific file
/claude-team-review path/to/plan.md

# Use maximum reasoning effort for the reviewer
/claude-team-review xhigh

Reviewer behavior

The reviewer uses an adversarial stance — it defaults to skepticism and tries to break confidence in the change. Each finding must answer:

  1. What can go wrong? — concrete scenario
  2. Why vulnerable? — cite specific location
  3. Impact — what breaks and how badly
  4. Recommendation — specific fix

The reviewer verifies findings by running tests, checking documentation, and inspecting related code before reporting.

Roadmap

  • Real-world testing and iteration on prompts
  • Parallel multi-reviewer mode (security + performance + correctness)
  • Persistent reviewer memory across sessions
  • Integration with CI (GitHub Actions)
  • Comparison benchmarks: Codex backend vs Team backend

Эксперимент: сравнение ревьюеров

Мы запустили оба ревьюера (Opus и GPT-5.4) на одном и том же плане и сравнили находки. Ключевой вывод: модели ревьюят из принципиально разных парадигм — Opus как архитектор ("сработает ли этот дизайн?"), Codex как security/ops инженер ("что сломается в продакшене?"). Ноль полных совпадений, ~30% частичных пересечений.

Подробности: EXPERIMENT.md — полный ход эксперимента, все находки, анализ пересечений, выводы.

License

Apache-2.0 — see LICENSE.

S
Description
No description provided
Readme
232 KiB
Languages
Markdown 100%