RAG · Search

It replaces flat vector dumps with a four-tier semantic pyramid and Mermaid symbol graphs so agents remember workflows without drowning in their own tool logs.

Memora is an MCP server that stores agent outputs in a durable, semantically searchable knowledge graph so AI assistants can recall context across sessions.

A Python layer that makes AI agents explain themselves through structured context graphs, decision trails, and W3C-compliant provenance.

A privacy-first Android fork that runs LLMs, image generation, and speech AI entirely offline, then locks itself behind your fingerprint.

SoloMD is a local-first markdown editor that bundles an MCP server so Claude, Cursor, or Codex can read and write your vault—while forcing every agent edit onto a branch you accept or reject.

It exists because builders wanted a Claude Cowork-style agent that keeps work context local, encrypted, and auditable instead of shipping memory to someone else’s server.

Arkon is a self-hosted knowledge hub that doesn't just chunk documents—it compiles them into a reviewed, versioned wiki and serves it to Claude through MCP.

Datus is an open-source agent that keeps LLMs from hallucinating SQL by building a living, learning knowledge base around your data stack.

Rocketnotes pairs a hierarchical note editor with an AI agent that reads your scattered snippets and archives them where they belong.

It turns your multi-language codebase into a Memgraph knowledge graph you can query in English—and then edit with surgical precision.

FalkorDB is a Redis module that uses sparse-matrix linear algebra to query property graphs fast enough for LLM retrieval workloads.

It turns unstructured text into an interactive knowledge graph by chunking documents, extracting Subject-Predicate-Object triplets with an LLM, and stitching the results together across chunks.

SIE replaces the usual tangle of separate embedding, reranking, and extraction servers with a single open-source container that scales from a laptop to Kubernetes.

Gortex indexes your code into a queryable knowledge graph so AI agents can ask for exactly the context they need instead of wasting tokens on full-file reads.

A self-hosted framework for building real-time voice-first AI agents that persist memory, delegate long tasks to background sub-agents, and optionally show up as lip-synced digital humans.

RAG Web UI bundles document ingestion, vector search, and multi-provider LLM dialogue into a single dockerized stack so you don't have to wire FastAPI and LangChain together for the hundredth time.

PipesHub exists so teams can search, automate, and chat across their enterprise data without surrendering it to a third-party cloud.

VideoRAG indexes absurdly long videos into a multi-modal knowledge graph so you can query them in natural language instead of scrubbing timelines.

ReMe gives forgetful AI agents long-term memory by storing context in readable files instead of opaque databases.

It packages Karpathy's LLM wiki concept as an Agent Skill so your coding agent can maintain a durable knowledge base with citations instead of re-deriving answers from raw sources on every query.