Awesome Context Engineering

v1.0.0Knowledge & Memorystable

A curated list of awesome open-source libraries for context engineering (Long-term memory, MCP: Model Context Protocol, Prompt/RAG Compression, Multi-Agent)

awesome-context-engineeringmcpai-integration
Share:
107
Stars
0
Downloads
0
Weekly
0/5

What is Awesome Context Engineering?

Awesome Context Engineering is a Model Context Protocol (MCP) server that allows AI assistants like Claude, Cursor, and VS Code to curated list of awesome open-source libraries for context engineering (long-term memory, mcp: model context protocol, prompt/rag compression, multi-agent)

A curated list of awesome open-source libraries for context engineering (Long-term memory, MCP: Model Context Protocol, Prompt/RAG Compression, Multi-Agent)

This server falls under the Knowledge & Memory category on MCPgee, the world's largest MCP server directory with 33,000+ servers.

Features

  • A curated list of awesome open-source libraries for context

Use Cases

Discover curated open-source libraries for context engineering, long-term memory, prompt compression, and multi-agent systems. Explore best practices in RAG and knowledge management.
jihoo-kim

Maintainer

LicenseMIT
Languagetypescript
Versionv1.0.0
UpdatedApr 9, 2026
Statushealthy
Maintenanceactive

Works with

ClaudeOpenAIwindowsmacoslinux

Installation

Manual Installation

npx awesome-context-engineering

Configuration

Configuration Details

Config File

claude_desktop_config.json

Performance

Response Metrics

Response Time< 200ms
ThroughputMedium

Resource Usage

Memory UsageLow
CPU UsageLow

How to Set Up and Use Awesome Context Engineering

The Awesome Context Engineering repository is a curated, community-maintained reference list of 40+ open-source libraries organized around the four core operations of context engineering for LLMs: writing long-term memory, selecting relevant context, compressing prompts and RAG outputs, and isolating context across multi-agent systems. It covers frameworks like mem0, Letta, and graphiti for persistent agent memory; FastMCP and MCP SDKs for Model Context Protocol integration; LLMLingua and xRAG for prompt and retrieval compression; and MetaGPT, agno, and camel for multi-agent orchestration. Developers use it as a starting point when architecting production AI systems that need to manage information efficiently within context windows.

Prerequisites

  • A GitHub account to star, fork, or contribute to the repository
  • Familiarity with Python or TypeScript (most listed libraries use one of these)
  • Understanding of basic LLM concepts: context window, RAG, agents
  • No installation required — this is a reference list, not an installable package
1

Browse the repository on GitHub

Visit the repository to explore all four categories of context engineering libraries. The README is the primary artifact — it links directly to each project.

# Visit: https://github.com/jihoo-kim/awesome-context-engineering
2

Identify the context operation you need

Determine which of the four pillars applies to your use case: Write (long-term memory for agents), Select (retrieval and MCP tools), Compress (reducing prompt size), or Isolate (multi-agent partitioning).

3

Explore memory libraries (Write Context)

For persistent agent memory, evaluate mem0 (managed memory service), Letta (stateful agent framework), graphiti (knowledge graph memory), and cognee (semantic memory). Each project README explains installation and usage.

# Example: install mem0
pip install mem0ai

# Example: install Letta
pip install letta
4

Explore MCP and retrieval tools (Select Context)

For dynamic context selection via Model Context Protocol, explore fastmcp (FastAPI-style MCP server builder) and the official mcp-python-sdk. These let you build servers that serve relevant context on demand.

# Example: install fastmcp
pip install fastmcp

# Example: install mcp SDK
pip install mcp
5

Explore compression libraries (Compress Context)

To reduce token costs without losing key information, review LLMLingua (token-level compression), sammo (semantic compression), xRAG (extreme RAG compression), and recomp (retrieval compressor).

# Example: install LLMLingua
pip install llmlingua
6

Explore multi-agent frameworks (Isolate Context)

For systems where specialized agents each receive only relevant context, review MetaGPT, agno, camel, and langroid. Each framework has its own architecture for partitioning information flow between agents.

# Example: install agno
pip install agno

Awesome Context Engineering Examples

Client configuration

Example of a FastMCP-based context server built using a library from the Awesome Context Engineering list.

{
  "mcpServers": {
    "context-server": {
      "command": "python",
      "args": ["-m", "your_fastmcp_server"],
      "env": {
        "MEMORY_BACKEND": "mem0",
        "MEM0_API_KEY": "your-mem0-api-key"
      }
    }
  }
}

Prompts to try

Questions to ask an AI assistant when researching context engineering patterns.

- "What are the best open-source libraries for adding long-term memory to an AI agent?"
- "Compare mem0 vs Letta for persistent agent memory — which fits a production use case better?"
- "How does LLMLingua compress prompts, and how much token reduction can I expect?"
- "What MCP libraries exist for building context-selection tools?"
- "Explain the difference between RAG compression and multi-agent context isolation"

Troubleshooting Awesome Context Engineering

A listed library link is broken or the project has been archived

Open an issue or submit a pull request on the awesome-context-engineering repository to report or fix broken links. The community maintains the list collaboratively.

Unsure which category of context engineering to use for a specific problem

Start with the problem statement: if your agent forgets things across sessions, use Write (memory). If it retrieves too much irrelevant text, use Select or Compress. If multiple agents interfere with each other's context, use Isolate.

A library in the list has incompatible dependency versions

Use a virtual environment (venv or conda) per library to isolate dependencies. Check each project's own issues tracker for known version conflicts before filing a new report.

Frequently Asked Questions about Awesome Context Engineering

What is Awesome Context Engineering?

Awesome Context Engineering is a Model Context Protocol (MCP) server that curated list of awesome open-source libraries for context engineering (long-term memory, mcp: model context protocol, prompt/rag compression, multi-agent) It connects AI assistants to external tools and data sources through a standardized interface.

How do I install Awesome Context Engineering?

Follow the installation instructions on the Awesome Context Engineering GitHub repository. Clone the repo, install dependencies, and add the server config to your AI client.

Which AI clients work with Awesome Context Engineering?

Awesome Context Engineering works with all major MCP-compatible AI clients including Claude Desktop, Claude Code, Cursor, VS Code (GitHub Copilot), Windsurf, and Cline.

Is Awesome Context Engineering free to use?

Yes, Awesome Context Engineering is open source and available under the MIT license. You can use it freely in both personal and commercial projects.

Browse More Knowledge & Memory MCP Servers

Explore all knowledge & memory servers available in the MCPgee directory. Each server includes setup guides for Claude, Cursor, and VS Code.

Quick Config Preview

{ "mcpServers": { "awesome-context-engineering": { "command": "npx", "args": ["-y", "awesome-context-engineering"] } } }

Add this to your claude_desktop_config.json or .cursor/mcp.json

Read the full setup guide →

Ready to use Awesome Context Engineering?

Browse our complete directory of 33,000+ MCP servers, read setup guides for your editor, and start building with the Model Context Protocol.

33,000+ ServersFree & Open SourceStep-by-Step Guides