help
Access documentation and guidance for the MCP Contemplation system, enabling understanding of its continuous cognitive processing and insight generation features.
Instructions
Get help documentation for contemplation system
Input Schema
TableJSON Schema
| Name | Required | Description | Default |
|---|---|---|---|
No arguments | |||
Implementation Reference
- src/index.ts:571-575 (handler)Handler function for the 'help' tool that returns the predefined HELP_DOCUMENTATION as a text content response.case 'help': { return { content: [{ type: 'text', text: HELP_DOCUMENTATION }], }; }
- src/index.ts:481-488 (registration)Registration of the 'help' tool in the ListTools response, specifying name, description, and empty input schema.{ name: 'help', description: 'Get help documentation for contemplation system', inputSchema: { type: 'object', properties: {}, }, },
- src/index.ts:484-487 (schema)Input schema definition for the 'help' tool, which takes no parameters.inputSchema: { type: 'object', properties: {}, },
- src/index.ts:13-78 (helper)Helper constant containing the full markdown documentation text returned by the 'help' tool.const HELP_DOCUMENTATION = ` # MCP Contemplation Server Interface to Claude's background contemplation loop - a persistent subprocess that: - Processes thoughts asynchronously between conversations - Works with local Ollama models for different thinking styles - Manages context to avoid overflow - Saves insights to both temporary scratch and permanent Obsidian storage - Learns from usage patterns to improve insight selection ## Available Functions: ### start_contemplation() Starts the background contemplation loop if not already running. Returns: Status message ### send_thought(thought_type, content, priority?) Sends a thought for background processing. - thought_type: "pattern", "connection", "question", or "general" - content: The thought content to process - priority: Optional priority (1-10, default 5) Returns: Thought ID for tracking ### get_insights(thought_type?, limit?, min_significance?) Retrieves processed insights from the contemplation loop. - thought_type: Optional filter by type - limit: Max number of insights (default 10) - min_significance: Minimum significance score (default 5) Returns: Array of insights with metadata ### set_threshold(significance_threshold) Sets the minimum significance for insights to be returned. - significance_threshold: Number 1-10 (default 5) Returns: Confirmation message ### get_memory_stats() Gets memory usage statistics for the contemplation system. Returns: Stats object with insight counts, memory usage, etc. ### get_status() Gets the current status of the contemplation loop. Returns: Status object with running state, queue size, etc. ### stop_contemplation() Gracefully stops the contemplation loop. Returns: Status message ### clear_scratch() Clears temporary scratch notes (keeps Obsidian permanent notes). Returns: Number of files cleared ### help() Returns this documentation. ## Thought Types: - **pattern**: Notice recurring themes across conversations - **connection**: Find links between disparate ideas - **question**: Explore interesting questions that arise - **general**: Open-ended reflection ## How It Works: The contemplation loop runs as a background process, using local LLMs (via Ollama) to process thoughts between conversations. High-significance insights are saved permanently to Obsidian, while medium-significance thoughts go to temporary scratch storage for 4 days. The system learns which insights prove valuable over time. `;