MCP Server: ai.stompy/stompy (Stompy)

The ai.stompy/stompy MCP server provides persistent AI memory for Claude and MCP agents. It stores, searches, and recalls project knowledge across conversations to reduce context loss. The system supports semantic retrieval, contradiction checking, document ingestion, and ticket tracking with per-ticket state machines.

🛠️ Key Features

• Persistent memory to prevent context loss between AI sessions
• Semantic Search using hybrid keyword + vector search over project contexts
• Conflict Detection using NLI-powered contradiction detection across stored knowledge
• Ticket Management with built-in ticketing and state machines per ticket type
• Document Ingestion with upload and AI chunking
• Multi-Project storage with isolated storage per project and cross-project search

🚀 Use Cases

• Remember project context across separate AI sessions
• Retrieve relevant knowledge using semantic search across stored contexts
• Detect contradictions in previously stored information
• Track work items through ticket state machines
• Ingest documents and split them into AI-processed chunks

⚡ Developer Benefits

• Persistent project knowledge storage across conversations
• Search and recall of project contexts via hybrid keyword + vector search
• Automated contradiction detection via NLI across stored knowledge
• Project isolation with optional cross-project search

⚠️ Limitations

• Described tool inventory is limited to 9 tools (toolCount: 9).