Repository navigation
FAQ
Context Processor is an MCP server that helps Claude and other AI agents manage, organize, and enhance information through intelligent pre-processing strategies.
Context Processor is designed for content enhancement with built-in analysis strategies. Databases are for data storage and retrieval. Think of Context Processor as a smart filing system with improvement capabilities.
Yes! It's a fully functional Node.js library usable standalone:
import { ContextProcessor } from "context-processor";
const processor = new ContextProcessor();
processor.save({ title: "...", content: "..." });Latest is 1.0.1+
npm install context-processor@latestNo! File-based storage is built-in. Contexts are stored as JSON files in a local directory.
Not yet, but it's planned for v1.1+. Currently only file-based storage is supported.
Use tags:
processor.save({
title: "React Guide",
content: "...",
tags: ["react", "frontend", "tutorial"]
});
processor.search({ tags: ["react"] });Also use metadata for classification:
processor.save({
title: "...",
content: "...",
metadata: {
category: "Frontend",
difficulty: "Beginner"
}
});| Tags | Metadata |
|---|---|
| For searching | For classification |
| Exact match | Flexible |
| Built-in search | You filter manually |
| Array of strings | Any object |
Practically: 10,000+ contexts work well. For 100,000+, consider an external database. Use pagination and batching for performance.
Yes, re-save with the same ID:
const updated = processor.save({
id: "ctx-123",
title: "Updated Title",
content: "New content..."
});- Strategies - Individual analysis techniques (Clarify, Analyze, etc.)
- Models - Pre-configured combinations of strategies
| Goal | Model |
|---|---|
| Improve clarity | clarify |
| Better search/SEO | search_optimized |
| Content metrics | analysis |
| Everything | comprehensive |
| URLs/links | web_enhanced |
| No processing | None (default) |
Ranked by speed (fastest to slowest):
- None (no preprocessing)
-
clarifyorsearch_optimized analysisweb_enhancedcomprehensive
Typical performance:
- Save (no model): ~50ms
- Save (with clarify): ~100ms
- Save (comprehensive): ~300-500ms
- Load: ~10-20ms
- List 100: ~50ms
- Search by tag: ~10ms
- Don't always use comprehensive - use specific models
- Batch operations together
- Use pagination for large lists
- Clean up old contexts regularly
- Monitor memory usage
- Install:
npm install context-processor - Configure as MCP server in Claude Code settings
- Claude will automatically have access to the 6 tools
Yes! Fully typed with interfaces and strict mode support.
Yes! File-based storage means:
- No cloud transmission
- Local storage only
- You control access
- No external dependencies
Security tips:
- Restrict file permissions:
chmod 700 ./contexts - Don't commit to public repos
- Use
.gitignorefor contexts directory
- Verify contexts exist:
processor.list() - Check tags match exactly (case-sensitive)
- Remember: Multiple tags use AND logic (all must match)
- Using
comprehensivemodel? Try a simpler one - Processing thousands? Use batches
- Memory limited? Increase Node memory:
node --max-old-space-size=4096 app.js
- Custom strategy API
- Database adapter support
- Performance improvements
- More built-in models
Absolutely! We welcome:
- Bug reports
- Feature requests
- Code contributions
- Documentation improvements
- Test scenarios
See Contributing Guide.
- This Wiki - Most answers are here
- Troubleshooting - Common issues
- GitHub Issues - Report bugs
- Discussions - Ask questions
- Community Testing - Join us!
Yes! Check:
- Getting Started - Quick examples
- Scenario 1 - Knowledge base example
- Scenario 2 - Analysis examples
- Scenario 3 - Advanced examples
Didn't find your answer? Open an issue or join Community Testing!