Atom-of-thoughts

Instructions

Implementation Reference

• src/index.ts:493-557 (handler)

  The core handler logic for the 'AoT-light' tool, implemented as the overridden processAtom method in AtomOfThoughtsLightServer class. This method validates input, stores atoms, handles verification and conclusion suggestions in a lightweight manner, and generates responses.
  public processAtom(input: unknown): { content: Array<{ type: string; text: string }>; isError?: boolean } { try { const validatedInput = this.validateAtomData(input); // Store the atom this.atoms[validatedInput.atomId] = validatedInput; // Add to order if it's new if (!this.atomOrder.includes(validatedInput.atomId)) { this.atomOrder.push(validatedInput.atomId); } // Format and display the atom with simplified output const formattedAtom = this.formatAtom(validatedInput); console.error(formattedAtom); // Quick verification - if verification atom, immediately verify dependencies if (validatedInput.atomType === 'verification' && validatedInput.isVerified) { validatedInput.dependencies.forEach(depId => { if (this.atoms[depId]) { this.verifyAtom(depId, true); } }); } // Faster conclusion suggestion - if hypothesis with high confidence, suggest conclusion immediately if (validatedInput.atomType === 'hypothesis' && validatedInput.confidence >= 0.8) { this.suggestConclusion(validatedInput); } // Simplified termination check const shouldTerminate = this.shouldTerminate(); const bestConclusion = shouldTerminate ? this.getBestConclusion() : null; // Basic response with less processing return { content: [{ type: "text", text: JSON.stringify({ atomId: validatedInput.atomId, atomType: validatedInput.atomType, isVerified: validatedInput.isVerified, confidence: validatedInput.confidence, atomsCount: Object.keys(this.atoms).length, bestConclusion: bestConclusion ? { atomId: bestConclusion.atomId, content: bestConclusion.content, confidence: bestConclusion.confidence } : null }, null, 2) }] }; } catch (error) { return { content: [{ type: "text", text: JSON.stringify({ error: error instanceof Error ? error.message : String(error), status: 'failed' }, null, 2) }], isError: true }; } }
• src/index.ts:650-710 (schema)

  The tool schema definition for 'AoT-light', including name, description, and input schema for atom processing.
  const AOT_LIGHT_TOOL: Tool = { name: "AoT-light", description: `A lightweight version of Atom of Thoughts (AoT) designed for faster processing and quicker results. This streamlined version sacrifices some depth of analysis for speed, making it ideal for time-sensitive reasoning tasks. When to use: - Quick brainstorming sessions requiring atomic thought organization - Time-sensitive problem solving where speed is prioritized over exhaustive analysis - Simpler reasoning tasks that don't require deep decomposition - Initial exploration before using the full AoT for deeper analysis - Learning or demonstration purposes where response time is important Key differences from full AoT: - Lower maximum depth (3 instead of 5) for faster processing - Simplified verification process - Immediate conclusion suggestion for high-confidence hypotheses - Reduced computational overhead and response payload - Optimized for speed rather than exhaustive analysis Atom types and parameters are the same as the full AoT tool.`, inputSchema: { type: "object", properties: { atomId: { type: "string", description: "Unique identifier for the atom" }, content: { type: "string", description: "Actual content of the atom" }, atomType: { type: "string", enum: ["premise", "reasoning", "hypothesis", "verification", "conclusion"], description: "Type of atom" }, dependencies: { type: "array", items: { type: "string" }, description: "List of IDs of other atoms this atom depends on" }, confidence: { type: "number", minimum: 0, maximum: 1, description: "Confidence level of this atom (value between 0-1)" }, isVerified: { type: "boolean", description: "Whether this atom has been verified" }, depth: { type: "number", description: "Depth level of this atom (optional, defaults to 0)" } }, required: ["atomId", "content", "atomType", "dependencies", "confidence"] } };
• src/index.ts:771-773 (registration)

  Registration of the 'AoT-light' tool in the MCP server's list of available tools.
  server.setRequestHandler(ListToolsRequestSchema, async () => ({ tools: [AOT_TOOL, AOT_LIGHT_TOOL, ATOM_COMMANDS_TOOL], }));
• src/index.ts:778-780 (registration)

  Dispatch registration in the CallToolRequest handler that routes calls to 'AoT-light' to the light server's processAtom method.
  } else if (request.params.name === "AoT-light") { return atomLightServer.processAtom(request.params.arguments); } else if (request.params.name === "atomcommands") {
• src/index.ts:486-558 (helper)

  The AtomOfThoughtsLightServer class that implements the lightweight AoT logic, extending the base server with reduced max depth and simplified processAtom.
  class AtomOfThoughtsLightServer extends AtomOfThoughtsServer { constructor() { // Lower max depth for faster processing super(3); } // Override to simplify the verification process public processAtom(input: unknown): { content: Array<{ type: string; text: string }>; isError?: boolean } { try { const validatedInput = this.validateAtomData(input); // Store the atom this.atoms[validatedInput.atomId] = validatedInput; // Add to order if it's new if (!this.atomOrder.includes(validatedInput.atomId)) { this.atomOrder.push(validatedInput.atomId); } // Format and display the atom with simplified output const formattedAtom = this.formatAtom(validatedInput); console.error(formattedAtom); // Quick verification - if verification atom, immediately verify dependencies if (validatedInput.atomType === 'verification' && validatedInput.isVerified) { validatedInput.dependencies.forEach(depId => { if (this.atoms[depId]) { this.verifyAtom(depId, true); } }); } // Faster conclusion suggestion - if hypothesis with high confidence, suggest conclusion immediately if (validatedInput.atomType === 'hypothesis' && validatedInput.confidence >= 0.8) { this.suggestConclusion(validatedInput); } // Simplified termination check const shouldTerminate = this.shouldTerminate(); const bestConclusion = shouldTerminate ? this.getBestConclusion() : null; // Basic response with less processing return { content: [{ type: "text", text: JSON.stringify({ atomId: validatedInput.atomId, atomType: validatedInput.atomType, isVerified: validatedInput.isVerified, confidence: validatedInput.confidence, atomsCount: Object.keys(this.atoms).length, bestConclusion: bestConclusion ? { atomId: bestConclusion.atomId, content: bestConclusion.content, confidence: bestConclusion.confidence } : null }, null, 2) }] }; } catch (error) { return { content: [{ type: "text", text: JSON.stringify({ error: error instanceof Error ? error.message : String(error), status: 'failed' }, null, 2) }], isError: true }; } } }