# /analyze - Deep Codebase Analysis

Scan the codebase to extract design preferences, coding conventions, and architectural patterns.{{#IF_MEMORY}} Stores findings in vector DB.{{/IF_MEMORY}}

## Usage

```bash
/analyze                    # Full analysis (all dimensions)
/analyze design-system      # Single dimension
/analyze component-patterns # Single dimension
```

## Dimensions

`component-patterns`, `design-system`, `state-management`, `error-handling`, `api-patterns`, `naming-conventions`, `typescript-patterns`, `animation-patterns`, `layout-patterns`, `performance-patterns`

## Instructions

### 1. Determine Scope

Full analysis or single dimension based on argument.

### 2. Collect Samples

For each dimension, read 3-6 representative files (first 150 lines each). Use appropriate glob patterns for the dimension.

### 3. Extract Patterns

For each dimension, extract:
- **Preferences**: Subjective style choices
- **Conventions**: Rules consistently followed
- **Patterns**: Recurring architectural approaches
- **Anti-patterns**: Things deliberately avoided

### 4. Display Results

```
{{RAQR_FRAME_START}}
{{RAQR_ART_THINKING}}
{{RAQR_FRAME_END}}

🦝 Raqr · /analyze                          Traqr · {{PROJECT_NAME}}
{{RAQR_HR}}

Deep analysis in progress...

{{RAQR_HR_MEDIUM}}
CODEBASE ANALYSIS
============================
Design System (N findings)
  [preference] <finding>
  [convention] <finding>

Component Patterns (N findings)
  [pattern] <finding>
============================
Total: N findings across M dimensions

{{RAQR_HR}}
What's next? → /capture to store specific findings, or /voice for writing style analysis
```

{{#IF_MEMORY}}### 5. Store Findings

For each finding:
```bash
curl -s -X POST "{{MEMORY_API_BASE}}/memory/store" \
  -H "Content-Type: application/json" \
  -d '{
    "content": "<finding>",
    "category": "pattern",
    "tags": ["codebase-analysis", "<dimension>"],
    "sourceType": "codebase_analysis",
    "sourceProject": "{{TRAQR_PROJECT_SLUG}}",
    "confidence": 0.7
  }'
```

Deduplicate: search before storing, skip if similarity > 0.85.
{{/IF_MEMORY}}

### Error Handling

If any step fails, show the error with Raqr:

```
{{RAQR_FRAME_START}}
{{RAQR_ART_ALERT}}
{{RAQR_FRAME_END}}

Raqr · /analyze                            Traqr · {{PROJECT_NAME}}
{{RAQR_HR}}

Something went wrong during analysis.

<translate the raw error to plain English using the table below>

{{RAQR_HR}}
```

| Raw Error | Plain English | Fix |
|-----------|--------------|-----|
| `ECONNREFUSED` on memory API | Can't reach the memory server | Check if the platform is running, or skip memory storage |
| `429 Too Many Requests` | Rate limited by the API | Wait a moment and retry |
| No files found for dimension | No matching source files in this project | Skip this dimension — not all projects use every pattern |
| `ENOENT` on file read | File was deleted or moved since scan started | Re-run the analysis |

If no patterns are found at all:

```
{{RAQR_FRAME_START}}
{{RAQR_ART_SAD}}
{{RAQR_FRAME_END}}

Raqr · /analyze                            Traqr · {{PROJECT_NAME}}
{{RAQR_HR}}

No patterns found. The project might be too small or too new.

Try: /analyze component-patterns (narrower scope often finds more)

{{RAQR_HR}}
```