widescreen-research: Command for Claude Code

Post-Edit Operation Hook

Description

This hook validates that MCP server operations follow the State Tracker Pattern - servers should only track state, never provide intelligence.

Trigger

Runs after editing operation files in MCP servers that implement state tracking patterns.

Instructions

1. Run Validation

# For projects with state tracker validation
npm run validate:state-tracker

2. Review Output

If violations are found:

❌ Errors (Must Fix)

• Content generation (generateIdeas, createSolution, etc.)
• Content analysis (analyzeQuality, evaluatePrompt, etc.)
• Intelligence provision (infer, deduce, conclude)

⚠️ Warnings (Review Carefully)

• Decision making based on content
• Pattern matching on prompt content
• Potential content transformation

3. Fix Violations

BAD: Server tries to be intelligent

// ❌ Generating content
const ideas = this.generateIdeas(prompt);

// ❌ Analyzing content
const quality = this.evaluateQuality(prompt);

// ❌ Making decisions
if (prompt.includes('complex')) {
  return this.useAdvancedMode();
}

GOOD: Server tracks state only

// ✅ Store what LLM provides
const idea = { content: prompt };

// ✅ Use LLM's evaluation
const quality = parameters.quality;

// ✅ Use LLM's decision
const mode = parameters.mode;

4. Commit Changes

Once validation passes:

git add src/tools/operations/...
git commit -m "description"

The git pre-commit hook will also run validation automatically.

Checklist

Before committing your changes:

• Validation passes (npm run validate:state-tracker)
• Server only validates parameters
• Server only stores state
• Server only returns progress metadata
• No content generation
• No content analysis
• No intelligent decisions
• LLM provides all intelligence via prompt/parameters

Quick Reference

What the Server Can Do

✅ Validate parameter types and requirements ✅ Store what the LLM sends (unmodified) ✅ Count/aggregate stored items ✅ Calculate progress percentages ✅ Group items by LLM-provided categories ✅ Find items by LLM-provided scores/ratings ✅ Format metadata for display