Headless, peer-to-peer context synchronization for local AI agents. It enables multiple LLM agents to share structured context and resolve state conflicts over a serverless P2P network.
MCP server implementing ReAct reasoning-acting loops with tools like search, lookup, and finish, supporting both fixture and OpenAI LLMs for offline-first, traceable agent execution.
MCP-native TypeScript implementation of the Reflexion verbal RL loop for language agents, providing tools to start tasks, act, evaluate, reflect, and retry with pluggable LLMs and memory.
Semantic caching MCP server for AI agent tool calls, providing exact and similarity-based cache lookup, store, invalidation, and metrics via MCP tools.
Evidence-weighted trust graph for multi-agent systems. Enables querying trust and gating high-risk actions based on time-decayed evidence and endorsements.
An MCP server that validates tool calls against JSON Schema, performs deterministic repair, redacts secrets, and maintains a hash-chained audit ledger.
MCP server implementing the Generative Agents memory architecture, enabling observation, retrieval, reflection, and planning for agent memory management.
MCP server implementing Chain-of-Verification to reduce LLM hallucinations by drafting answers, planning fact-check questions, answering them independently, and revising.
Enables chain-of-thought reasoning with self-consistency voting, allowing sampling of diverse reasoning paths and aggregating answers via majority or weighted vote for improved accuracy.
Enables AI agents to fork plans into counterfactual worlds, score them with rubrics and simulations, detect contradictions, measure regret, and merge a winner with a full audit trail.
Enables AI agents to branch, score, prune, and merge conversation strands with provenance and timeline export. Provides MCP tools and resources for managing parallel exploratory strands.