Skill

investigate

Investigate code structures (/investigate) using a 3-phase process: Grep to locate, Read to verify, and Codex (optional) to cross-check. Prevents false reports from guessing. Covers all code structures: function calls, hashes, conditionals, method chains, class definitions, and configuration values. Activates on "調査", "investigate", "コード調査", "grep で調べて", "呼び出し箇所", "使われ方", "一覧", "全件", "どこで使われている".

From rite

Install

Run in your terminal

npx claudepluginhub b16b1rd/cc-rite-workflow --plugin rite

Tool Access

This skill uses the workspace's default tool permissions.

Skill Content

Similar Skills

skill-lookup

Searches, retrieves, and installs Agent Skills from prompts.chat registry using MCP tools like search_skills and get_skill. Activates for finding skills, browsing catalogs, or extending Claude.

prompts.chat

157.5k

prompt-lookup

Searches prompts.chat for AI prompt templates by keyword or category, retrieves by ID with variable handling, and improves prompts via AI. Use for discovering or enhancing prompts.

prompts.chat

157.5k

agent-eval

Compares coding agents like Claude Code and Aider on custom YAML-defined codebase tasks using git worktrees, measuring pass rate, cost, time, and consistency.

ecc

140.3k

Stats

Parent Repo Stars0

Parent Repo Forks0

Last CommitMar 27, 2026

Actions

View Source View Plugin View on GitHub View README

investigate

From rite

Context

When activated, this skill provides:

Structured Investigation Process

Phase 1: Scope clarification
Phase 2: Search (Grep)
Phase 3: Verification (Read)
Phase 4: Dynamic pattern tracking
Phase 5: Result aggregation
Phase 6: Cross-check (Codex or alternative)
Phase 7: Final report

Code Structure Coverage

Function calls and arguments
Hash/dictionary structures
Conditional expressions and branches
Method chains
Class/module definitions
Configuration values

Verification Guarantees

Every grep hit is verified by Read
Completeness check (grep hit count = Read verification count)
Cross-check via Codex MCP or Claude self-verification
No guessing — unverified values reported as "unverified"

Investigate Skill

ソースコードを正確に調査し、検証済みの結果を報告するスキル。

Auto-Activation Keywords

調査, investigate, コード調査
grep で調べて, 呼び出し箇所, 使われ方
一覧, 全件, どこで使われている
引数, パラメータ, 設定値
コード構造, パターン検索

Context

When activated, this skill provides:

Structured Investigation Process
- Phase 1: Scope clarification
- Phase 2: Search (Grep)
- Phase 3: Verification (Read)
- Phase 4: Dynamic pattern tracking
- Phase 5: Result aggregation
- Phase 6: Cross-check (Codex or alternative)
- Phase 7: Final report
Code Structure Coverage
- Function calls and arguments
- Hash/dictionary structures
- Conditional expressions and branches
- Method chains
- Class/module definitions
- Configuration values
Verification Guarantees
- Every grep hit is verified by Read
- Completeness check (grep hit count = Read verification count)
- Cross-check via Codex MCP or Claude self-verification
- No guessing — unverified values reported as "unverified"

Configuration

Reads investigate section from rite-config.yml:

investigate:
  codex_review:
    enabled: true    # Set false to skip Codex cross-check in Phase 6

When codex_review.enabled is false or Codex MCP is unavailable, Claude performs self-verification as an alternative (Phase 6b: 代替検証).

Command

Full investigation procedure: commands/investigate.md

Lightweight protocol for use in other phases: references/investigation-protocol.md