M
MCP Directory CURATED DIRECTORY
911 MCP servers indexed
中文
← Back to directory
L

LLM Context

Community
Smart context management for LLM development workflows.
How to install ↓ Compare GitHub source repository ↗
Category
Dev Tools #219 of 438
Stars
★ 307 Very popular
Transport
stdio (local process)
Runtime
Python
Credentials
No credential needed
License
Apache-2.0
Last commit
Tools
3
53FMRS · D

LLM Context is a practical MCP server focused on simplifying code sharing through rule-driven context management. It offers flexible customization and intelligent file selection, making it ideal for integration with MCP-capable tools like Claude Desktop. Community-maintained with quality code, but users should be mindful of permissions and potential sensitive information risks.

Strongest · Documentation 12/20 Weakest · Reliability 8/20

Reliability
8/20
Security and permissions
10/20
Maintenance
12/20
Documentation
12/20
Setup experience
11/20
Why each score
Reliability 8/20
Static analysis shows a well-structured CLI and MCP tool definitions, but no execution verification. The README's claimed tools align with code, yet error handling and edge cases lack test evidence. Dependency management uses uv, but no CI workflows or committed tests for the server's key paths were found, thus lower score.
Security and permissions 10/20
The project primarily handles local files with no obvious remote network requests or sensitive credential handling. MCP tools don't involve external API keys. However, permission and confirmation mechanisms are unclear, e.g., lc-missing may read arbitrary files without apparent path restrictions or confirmation. No malicious code or gross overreach detected, hence a medium-high score.
Maintenance 12/20
Repository is active with recent commits and releases, but no explicit maintenance policy or security response channel. License is Apache-2.0 and clear. A few open issues exist. Dependencies may not be updated promptly. Overall, good maintenance but lacks systematic governance.
Documentation 12/20
README and docs are very detailed, covering installation, usage, rule creation, and MCP configuration with examples. However, some advanced features (like path format) might need deeper documentation, and troubleshooting section is missing. Also, examples may not fully reflect the latest code, with hidden assumptions.
Setup experience 11/20
Installation steps are clear, with CLI and MCP config examples, quick install via uv tool. However, platform compatibility is not detailed, and it relies on external tool uvx. No explicit support for Windows, and setup may encounter issues in certain environments.

Static review · not runListed 2026-08-07

Read the FMRS scoring method →

Fit and risk

What it can accessReads local files

Best for

  • Developers using LLM-assisted development with Claude Desktop or other MCP-supported chat interfaces.
  • Teams needing to provide precise project context to AI agents.
  • Those who want to reduce manual copy-paste and optimize token usage.

Not for

  • Users who don't need to share code with LLMs.
  • Users in environments without MCP support (though CLI is still available).
  • High-security production requiring official maintenance (this is community project).

Required permissions

  • Read project files: requires access to project directory to select and read files.
  • Execute command-line operations: needs shell access for `lc-*` commands.
  • Optional clipboard access: `lc-context` copies to clipboard by default.

Risks and side effects

  • Sensitive data leakage: shared files may contain sensitive info; choose carefully.
  • Dependency on third-party tools: requires `uvx` or `pip`; may be affected by network.
  • Misconfiguration: incorrect rules may lead to oversized or missing context.

Setup

Before you start

Runtime：Python

Install via pip: pip install llm-context. Then add the MCP server to Claude Desktop configuration: in claude_desktop_config.json, set an mcpServers entry with command uvx and args --from llm-context lc-mcp.

claude_desktop_config.json 
{
  "mcpServers": {
    "llm-context": {
      "command": "uvx",
      "args": [
        "--from",
        "llm-context",
        "lc-mcp"
      ]
    }
  }
}

Shown for Claude Desktop. Other clients may use a different file or key (VS Code uses "servers") — the configurator below converts it.

.vscode/mcp.json
{
  "servers": {
    "llm-context": {
      "command": "uvx",
      "args": [
        "--from",
        "llm-context",
        "lc-mcp"
      ]
    }
  }
}

Goes in your project's .vscode/mcp.json (VS Code uses a "servers" key).

Terminal
claude mcp add llm-context -- uvx --from llm-context lc-mcp

Run it in a terminal; replace any <…> placeholders with your own values first.

Generate and validate client config

Check that it works

Add llm-context to your Claude Desktop config and restart; the tool list should show lc_outlines, lc_preview, and lc_missing. Alternatively, run uvx --from llm-context lc-mcp to confirm the server starts.

Troubleshooting

  1. Installation: ensure `llm-context` is properly installed and `lc-mcp` command is available.
  2. MCP configuration: verify command and args in `claude_desktop_config.json` are correct.
  3. Logs: check Claude Desktop logs for MCP connection errors.
  4. CLI test: try `lc-select` or `lc-context` to see if they work.

Things to try

Once connected, you can ask your AI assistant things like:

  • Generate code structure excerpts for the current project using lc_outlines
  • Use lc_preview to validate what files the tmp-prm-auth rule selects
  • Fetch the full contents of /src/auth.py with lc_missing
  • Switch to the prm-code rule and generate context for a code review

Tools 3

lc_outlines read-only
Generate excerpted context from current rule.
lc_preview read-only
Validate rule effectiveness (file selection and size) before use.
lc_missing read-only
Fetch specific files or implementations on demand.

Use cases

Quickly share relevant project files to LLM in chat interfaces.
Create reusable context rules for specific tasks (e.g., code review, documentation).
Let AI agents fetch additional files automatically via MCP or CLI.

Supported clients

Claude Desktop

Listed from the project's documentation, not tested by this site.

Overview

LLM Context is a community-maintained MCP server that shares relevant project files with large language models via stdio, using rule-based customization. It enables intelligent file selection, composable rules, code excerpting, and on-demand file retrieval through MCP tools (lc_outlines, lc_preview, lc_missing). It is developed by cyberchitta and licensed under Apache-2.0.

Similar servers

Context7 80 · B

Upstash's official server providing up-to-date third-party library docs for AI coding assistants

★ 63.3k · Tools 2 Compare with this →
Compare with other MCP servers

Source revision 589e31f2b389 Data synced 2026-10-19 Read the FMRS scoring method