adversarial-code-review
Adversarial Code Review
Core Principle: Review as if you're trying to break the code. Deliberately adopt hostile perspectives—each reveals issues the others miss.
This is not about finding fault. It's about finding problems before users do.
Review Mode
| Mode | Trigger | Focus |
|---|---|---|
| Diff-Focused (default) | No explicit instruction, PR review | What changed? What could break? |
| Audit | "audit", "holistic", "codebase review" | Broader scope, systematic coverage |
When in doubt, use diff-focused mode. Audit mode requires explicit request.
The Six Adversarial Lenses
Review through each lens deliberately. Don't blend them—switching perspectives forces deeper analysis.
More from third774/dotfiles
opensrc
Fetch source code for npm, PyPI, or crates.io packages and GitHub/GitLab repos to provide AI agents with implementation context beyond types and docs. Use when needing to understand how a library works internally, debug dependency issues, or explore package implementations.
90natural-writing
Write like a human, not a language model. Avoid AI-tell vocabulary, formulaic structures, and hollow emphasis. Apply to ALL written output including prose, documentation, comments, and communication. Use when drafting prose, documentation, comments, or any written output that should sound human.
66agent-skills
Author and improve Agent Skills following the agentskills.io specification. Use when creating new SKILL.md files, modifying existing skills, reviewing skill quality, or organizing skill directories with proper naming, descriptions, and progressive disclosure.
31documenting-code-comments
Standards for writing self-documenting code and best practices for when to write (and avoid) code comments. Use when auditing, cleaning up, or improving inline code documentation.
28customizing-opencode
Configure OpenCode via opencode.json, agents, commands, MCP servers, custom tools, plugins, themes, keybinds, and permissions. Use when setting up or modifying OpenCode configuration.
23verification-before-completion
Run verification commands before claiming work is complete or fixed. Use before asserting any task is done, bug is fixed, tests pass, or feature works.
18