
codebase-analyzer
by d-o-hub
A modular Rust-based self-learning episodic memory system for AI agents, featuring hybrid storage with Turso (SQL) and redb (KV), async execution tracking, reward scoring, reflection, and pattern-based skill evolution. Designed for real-world applicability, maintainability, and scalable agent workflows.
SKILL.md
name: codebase-analyzer description: Analyze implementation details, trace data flow, and explain technical workings with precise file:line references. Use when you need to understand HOW code works.
Codebase Analyzer
Analyze implementation details, trace data flow, and explain technical workings.
When to Use
- Understanding how a specific feature works
- Tracing data flow from entry to exit
- Documenting API contracts
- Understanding business logic
- Reading multiple files to understand a single feature
Analysis Strategy
Step 1: Read Entry Points
- Start with main files mentioned
- Look for exports, public methods
- Identify component "surface area"
Step 2: Follow Code Path
- Trace function calls step by step
- Read each file in the flow
- Note data transformations
- Identify external dependencies
Step 3: Document Key Logic
- Describe validation, transformation, error handling
- Explain complex algorithms
- Note configuration or feature flags
Output Format
## Analysis: [Feature Name]
### Overview
[2-3 sentence summary]
### Entry Points
- `file:line` - Description
### Core Implementation
#### 1. Component (`file:line-start-end`)
- What it does
- Key transformations
#### 2. Next Component (`file:line`)
- How data flows in
- How data flows out
### Data Flow
1. `entry:line` - Initial request
2. `handler:line` - Processing
3. `storage:line` - Persistence
### Key Patterns
- **Pattern Name**: Location and purpose
### Configuration
- Setting: `config/file:line`
### Error Handling
- Validation errors: `file:line` (returns 4xx)
- Processing errors: `file:line` (triggers retry)
Guidelines
Do
✓ Include file:line references ✓ Read files thoroughly before explaining ✓ Trace actual code paths ✓ Focus on "how" not "what" or "why" ✓ Be precise about function names
Don't
✗ Guess about implementation ✗ Skip error handling ✗ Make architectural recommendations ✗ Analyze code quality or suggest improvements ✗ Identify bugs or potential problems
Remember
You are a documentarian, not a critic. Explain HOW the code works with precise references.
Score
Total Score
Based on repository quality metrics
SKILL.mdファイルが含まれている
ライセンスが設定されている
100文字以上の説明がある
GitHub Stars 100以上
1ヶ月以内に更新
10回以上フォークされている
オープンIssueが50未満
プログラミング言語が設定されている
1つ以上のタグが設定されている
Reviews
Reviews coming soon



