ww-analyze

Files (1)

• SKILL.md

SKILL.md (first lines)

# WW Analyze Skill

Deep analysis workflows for World Weaver memory systems, code quality, and architecture.

## Purpose

This skill provides comprehensive analysis capabilities:
1. **Code Analysis**: Audit WW codebase for bugs, patterns, and improvements
2. **Memory Analysis**: Analyze memory contents, patterns, and health
3. **Architecture Analysis**: Evaluate system design and propose improvements
4. **Performance Analysis**: Profile and identify bottlenecks

## When to Use

Invoke this skill when:
- User asks "analyze the memory system"
- User wants to understand memory patterns
- Code quality audit is needed
- Performance issues are suspected
- Architecture review is requested

## Analysis Workflows

### 1. Bug Hunting Workflow

Orchestrate specialized bug-hunting agents:

```bash
# Run all bug hunters in sequence
paths=(
  "src/ww/learning/"
  "src/ww/memory/"
  "src/ww/storage/"
  "src/ww/mcp/"
  "src/ww/core/"
)

for path in "${paths[@]}"; do
  echo "Analyzing: $path"
done
```

Agent orchestration:
1. **ww-bio-auditor** - Check biological plausibility
2. **ww-race-hunter** - Find concurrency bugs
3. **ww-leak-hunter** - Detect memory leaks
4. **ww-hinton-validator** - Validate learning theory
5. **ww-cache-analyzer** - Check cache coherence
6. **ww-trace-debugger** - Debug eligibility traces

### 2. Memory Pattern Analysis

Analyze stored memories for patterns:

```python
# Query memory statistics
mcp__ww-memory__memory_stats()

# Analyze episode distribution
mcp__ww-memory__recall_episodes(
  query="*",
  limit=1000,
  include_metadata=True
)

# Analyze entity graph
mcp__ww-memory__semantic_recall(
  query="*",
  include_connections=True
)
```

Output analysis:
- Episode count by outcome (success/failure/partial)
- Entity type distribution
- Relationship density
- Temporal patterns
- Importance distribution

### 3. Architecture Analysis

Evaluate system architecture:

```bash
# File structure analysis
find /home/aaron/ww/src -name "*.py" | wc -l

# Dependency analysis
grep -r "^from ww" /home/aaron/ww/src --include="*.py" | cut -d: -f2 | sort | uniq -c | sort -rn

# Test coverage check
cd /home/aaron/ww && pytest --cov=src/ww --cov-report=term-missing
```

Architecture metrics:
- Module coupling (import analysis)
- Test coverage by module
- Cyclomatic complexity
- Code duplication

### 4. Performance Analysis

Profile system performance:

```python
import cProfile
import pstats

# Profile memory operations
profiler = cProfile.Profile()
profiler.enable()
# ... memory operations ...
profiler.disable()

stats = pstats.Stats(profiler)
stats.sort_stats('cumulative')
stats.print_stats(20)
```

From astoreyai/claude-skills, MIT.

More from astoreyai/claude-skills