{
  "markdown": "# grill\n\n[![Validated by NLPM](https://img.shields.io/endpoint?url=https://raw.githubusercontent.com/xiaolai/grill-for-claude/main/nlpm-badge.json)](https://github.com/xiaolai/grill-for-claude/blob/main/nlpm-badge.json)\n\nDeep codebase interrogation — multi-angle architecture review that forces rigorous, actionable analysis. **Works in both Claude Code and OpenAI Codex CLI.**\n\n## What it does\n\nGrill launches 5-6 specialized analysis units in parallel to interrogate your codebase from every angle: architecture, error handling, security, testing, and edge cases. You pick a review style, optionally add pressure tests, and get a comprehensive report with a prioritized fixing plan.\n\n- **6 specialized analysis units**: recon, architecture, error-handling, security, testing, edge-cases (shipped as agents under Claude Code, as skills under Codex)\n- **5 review styles**: Architecture Review, Hard-Nosed Critique, Multi-Perspective Panel, ADR Style, Paranoid Mode — plus a \"Select All\" option\n- **8 add-on pressure tests**: scale stress, hidden costs, principle violations, strangler fig migration, success metrics, before/after diagrams, assumptions audit, compact & optimize\n- **Evidence-based findings**: every finding requires file paths, line references, severity tags, and effort estimates\n- **Fixing plan**: phased action plan (Critical > High > Medium > Low) with dependency graph and effort totals\n\nPart of the [xiaolai plugin marketplace](https://github.com/xiaolai/claude-plugin-marketplace).\n\n## Installation\n\n### Claude Code\n\nTwo install paths — both reach the same code. Pick one:\n\n**Via Anthropic's official community marketplace** (curated; updates lag the maintainer's marketplace by up to ~24h):\n\n```\n/plugin marketplace add anthropics/claude-plugins-community\n/plugin install grill@claude-community\n```\n\n**Via the xiaolai marketplace** (latest version lands here first):\n\n```\n/plugin marketplace add xiaolai/claude-plugin-marketplace\n/plugin install grill@xiaolai\n```\n\n> **Install fails with \"Plugin not found in marketplace 'xiaolai'\"?** Your local marketplace clone is stale. Run `claude plugin marketplace update xiaolai` and retry — `plugin install` does not auto-refresh. (The community marketplace doesn't have this caveat.)\n\n| Scope | Command (xiaolai) | Effect |\n|-------|--------------------|--------|\n| **User** (default) | `/plugin install grill@xiaolai` | Available in all your projects |\n| **Project** | `/plugin install grill@xiaolai --scope project` | Shared with team via `.claude/settings.json` |\n| **Local** | `/plugin install grill@xiaolai --scope local` | Only you, only this repo |\n\nThe same `--scope` flags work with the `@claude-community` variant.\n\n### OpenAI Codex CLI\n\nCodex installs plugins from a marketplace via an **in-session TUI**, not a shell command. There is no `codex plugin install` verb.\n\n**One-time setup** (shell):\n\n```bash\ncodex plugin marketplace add xiaolai/claude-plugin-marketplace\n```\n\n**Per-plugin install** (inside a Codex session):\n\n1. Start a session: `codex` (in your project directory)\n2. Type `/plugins` — opens the plugin TUI\n3. Find grill in the `xiaolai` marketplace, install/enable it\n\n## Usage\n\n### In Claude Code: slash command\n\n```\n/grill:roast\n```\n\nOptionally target a subdirectory: `/grill:roast src/`\n\n### In OpenAI Codex CLI: skill invocation\n\nCodex uses the `$skill-name` syntax, not slash commands. Type either:\n\n```\n$grill-roast\n```\n\nOr just describe the task in natural language (\"do a multi-angle audit of this codebase\") and Codex's auto-match will load the skill from its description.\n\n> **Why the difference?** Claude Code maps a plugin's `commands/*.md` files to slash commands like `/grill:roast`. Codex deprecated custom prompts in favor of skills (description-matched on intent), so the same orchestration is shipped as a skill under Codex's layout.\n\n## How it works\n\n1. **Recon** — the recon unit surveys the codebase (structure, stack, size)\n2. **Choose review style** — pick from 5 styles + optional add-on pressure tests\n3. **Deep dive** — 4-5 analysis units run, each analyzing their specialty area (parallel under Claude Code, sequential under Codex)\n4. **Synthesis** — findings are deduplicated, attributed, and formatted per the chosen style\n5. **Report** — a markdown report with executive summary and phased fixing plan is saved to the target directory\n\n## Analysis units\n\n| Unit | Focus | Claude name | Codex name |\n|------|-------|-------------|------------|\n| Recon | Quick codebase survey (structure, stack, dependencies) | `recon` agent | `$grill-recon` skill |\n| Architecture | Core architecture analysis | `architecture` agent | `$grill-architecture` skill |\n| Error handling | Error handling & observability | `error-handling` agent | `$grill-error-handling` skill |\n| Security | Security surface analysis | `security` agent | `$grill-security` skill |\n| Testing | Testing & CI/CD analysis | `testing` agent | `$grill-testing` skill |\n| Edge cases | Race conditions, boundary values, failure modes (Paranoid Mode) | `edge-cases` agent | `$grill-edge-cases` skill |\n\nAll units share the `grill-core` skill, which defines severity tags, finding format, evidence standards, and the untrusted-input rule.\n\n## Severity tags\n\n- `[CRITICAL]` — Security vulnerability, data loss risk, correctness bug\n- `[HIGH]` — Significant reliability, maintainability, or performance impact\n- `[MEDIUM]` — Noticeable quality issue, not urgent\n- `[LOW]` — Minor improvement, address when touching the file\n- `[GOOD]` — Positive finding worth calling out\n\n## License\n\nISC\n",
  "bytes": 5563,
  "sha": "a8b407afc0c170909749e4cd929ae43512f70f70030f3327671ffa1cde8d94f7",
  "repo_slug": "xiaolai/grill-for-claude",
  "fonte": "repo",
  "truncated": false,
  "api": "https://agentalog.com/api/listings/plg_xiaolai_grill_for_claude_grill_5c3357b0/readme"
}