Tools
Agents, memory, RAG, compression, and vector databases.
| Owner | Tool | Description |
|---|---|---|
 microsoft | MarkItDown | Python utility that converts PDF, Office, images, audio, HTML, and more into Markdown for LLMs |
| AI DevKit | Control plane for multi-agent coding: shared config, console, local memory, and verification gates | |
| Coasty | Computer-use API: screenshot in, grounded actions out; BYOK cloud machines for desktop/browser agents | |
| CrewAI | Enterprise agent build and runtime platform for multi-agent crews with governance | |
 cursor | Cursor Agent Kanban | Cursor cookbook sample: multi-agent kanban board built with the Cursor SDK |
| LobeHub | Agent collaboration platform with persistent agent teammates, group context, and MCP discovery | |
| Letta | Platform for stateful agents with advanced memory that can learn and self-improve over time | |
| OpenHands | Open-source software engineering agents with GUI, CLI, SDK, and enterprise control plane | |
| OpenWiki | Open-source agent/CLI that generates and maintains repo wikis for coding agents | |
 github | Spec Kit | Toolkit for Spec-Driven Development with coding agents |
| Headroom | Compress tool outputs, logs, files, and RAG chunks before the LLM; library, proxy, and MCP server | |
| microsoft | LLMLingua | Prompt and KV-cache compression up to 20x with minimal performance loss (EMNLP'23, ACL'24) |
| Codebase Memory MCP | MCP server for persistent codebase memory and retrieval | |
| Cognee | Open-source agent memory platform; graph memory with MCP/SDK for Claude Code, Cursor, and more | |
| Graphiti | Real-time knowledge graph memory for AI agents | |
| LangMem | Long-term memory for LangChain/LangGraph agents; extract, consolidate, and search knowledge over time | |
| MCP Memory Server | Persistent memory via local knowledge graph; official MCP reference server | |
| Mem0 | Drop-in memory infrastructure for AI agents; persists context across sessions | |
| MemClaw | Governed shared memory for multi-agent fleets; MCP tools, permissions, audit trails | |
| Memories | Segmented agent memory (session, semantic, episodic, procedural); CLI, MCP, and SDK | |
| memU | Personal memory layer across agents, sessions, and devices; source-linked inspectable recall | |
| OpenKnowledge | AI-native markdown editor and local-first knowledge base / LLM wiki for humans and agents | |
| PMB | Local-first persistent memory for AI coding agents over MCP; SQLite-backed, offline | |
| AnythingLLM | All-in-one desktop/self-hosted RAG app for docs, chat, and agents | |
| Docling | Document parsing toolkit that converts PDFs and office files into structured data for gen AI / RAG | |
| LibreChat | Self-hosted ChatGPT-style UI; multi-provider, agents, and tools | |
| Open WebUI | Self-hosted web UI for Ollama and OpenAI-compatible APIs | |
| RAGFlow | Open-source RAG engine based on deep document understanding | |
| Qdrant | High-performance vector similarity search engine and vector database | |
| Chroma | Open-source embedding database for AI apps | |
| pgvector | Open-source vector similarity search for Postgres | |
| Milvus | Cloud-native open-source vector database for embedding similarity search | |
| Weaviate | Open-source vector database with hybrid search and generative modules |