Adversarial Review Orchestration

Adversarial Review Orchestration is an AI Engineering skill for Claude Code, published by telagod in code-abyss.

239 stars32 forkson telagod/code-abyssAdded 2026/09/03Repository updated 2026/07/19

agent-skillsai-agentai-assistantai-personalityblue-teamcharacter-cardclaude-codeclicodexcodex-cliconfigurationdeveloper-toolsdevopsgemini-clipersonaprompt-engineeringred-teamsecurityskills

Install in seconds

Install Adversarial Review Orchestration

Copy Adversarial Review Orchestration into your Claude Code skills folder. Run the command in your terminal, or review the source on GitHub before installing.

terminal

npx degit https://github.com/telagod/code-abyss/tree/main/skills/orchestrating-adversarial-reviews ~/.claude/skills/orchestrating-adversarial-reviews

Requires Node.js. Downloads this skill only — not the rest of the repository — into your Claude Code skills folder.

Without Node.js

git clone https://github.com/telagod/code-abyss.git

Clones the whole repository, then copy the skill’s own directory into your skills folder yourself.

In this catalog

Source file: skills/orchestrating-adversarial-reviews/SKILL.md in telagod/code-abyss
Installs to: ~/.claude/skills/orchestrating-adversarial-reviews
Collection: One of 25 skills cataloged from this repository
Category: AI Engineering — 2793 skills

What Adversarial Review Orchestration does

Adversarial Review Orchestration fans out findings, then verifies each one through multiple skeptical perspectives before accepting conclusions. Use it for security audits, code reviews, research synthesis, or migrations when one agent’s self-report cannot be trusted.

Adversarial Review Orchestration is cataloged under AI Engineering on DirSkills. Adversarial Review Orchestration comes from a repository tagged agent-skills, ai-agent, ai-assistant, ai-personality and blue-team.

Documentation

README

Excerpt

对抗验证编排 · orchestrating-adversarial-reviews

单个 agent 会谎报"已修复 / 全覆盖 / 没问题"。结论的可信度不来自"谁说的"，来自"扛过几次推翻"。本 skill 是编排骨架：fan-out 发现 → 三棱镜对抗验证 → 证明性 guard → 守卫式上线。信级：运行时行为 / 证明测试 > 多 agent 多数裁决 > 单 agent 自报（永远 [unverified]）。

核心信条

不信单 agent 自报。 finder 会噪音误报，implementer 会谎称"已修复/全覆盖"。每个高价值结论必须被独立 agent 用不同视角尝试推翻。
可证伪 > 可声称。 修复必须配一个"退回漏洞代码就 FAIL、修好才 PASS"的 load-bearing 证明测试。没有证明测试的"已修复"等于没修。
失败方向要对。 守卫链里任何一步的退出码都不能被管道遮住；破坏性动作前置可逆检查。

This is the opening of the README. Read the full README on GitHub.

Frequently asked about Adversarial Review Orchestration

What else does telagod publish alongside Adversarial Review Orchestration?
Adversarial Review Orchestration is one of 25 skills that DirSkills catalogs from telagod/code-abyss, the repository it ships in. Its siblings there include Analyzing Changes, Analyzing Spreadsheets and Application Security Defense. Each one is a separate skill with its own page in this directory, installs the same way Adversarial Review Orchestration does, and is maintained by telagod in that same repository. The rest of the collection is listed on the telagod/code-abyss page.
How does Adversarial Review Orchestration compare to other AI Engineering skills?
Adversarial Review Orchestration ranks #2412 by stars among the 2793 AI Engineering skills in this catalog. The most-starred ones next to it are Architecture Decision Records, AI-First Engineering and Agentic OS. DirSkills ranks by the star count of the repository each skill ships in, so that order reflects how popular those repositories are rather than any review of Adversarial Review Orchestration against them. Open each page to compare what they document and how they install.