mcp-tldr
# mcp-tldr
MCP server that exposes https://github.com/parcadei/llm-tldr code analysis tools to Claude Code (or any MCP client).
## Prerequisites
- Python 3.11+
- `tldr` CLI installed and in your PATH ([install instructions](https://github.com/parcadei/llm-tldr))
## Setup
```bash
git clone https://github.com/YOUR_USER/mcp-tldr.git
cd mcp-tldr
python3 -m venv .venv
.venv/bin/pip install -e .
```
## Add to Claude Code
Add this to your `~/.claude.json` under `mcpServers`:
```json
{
"mcpServers": {
"tldr": {
"type": "stdio",
"command": "/absolute/path/to/mcp-tldr/.venv/bin/python",
"args": ["server.py"],
"cwd": "/absolute/path/to/mcp-tldr"
}
}
}
```
Replace `/absolute/path/to/mcp-tldr` with the actual path where you cloned the repo.
Alternatively, if `tldr` is not in your PATH, set the `TLDR_BIN` environment variable:
```json
{
"tldr": {
"type": "stdio",
"command": "/absolute/path/to/mcp-tldr/.venv/bin/python",
"args": ["server.py"],
"cwd": "/absolute/path/to/mcp-tldr",
"env": {
"TLDR_BIN": "/path/to/tldr"
}
}
}
```
## Available Tools
| Tool | Description |
|------|-------------|
| `tldr_tree` | File tree of a project |
| `tldr_structure` | Code structure (classes, functions, imports) |
| `tldr_context` | LLM-ready context for a function/class |
| `tldr_search` | Pattern search across files |
| `tldr_impact` | Find all callers of a function |
| `tldr_calls` | Cross-file call graph |
| `tldr_semantic` | Natural language code search |
| `tldr_arch` | Detect architectural layers |
| `tldr_warm` | Pre-build cache for faster queries |
| `tldr_extract` | Full file analysis (classes, functions, methods) |
| `tldr_cfg` | Control flow graph for a function |
| `tldr_dfg` | Data flow graph for a function |
| `tldr_slice` | Program slice (what affects line X) |
| `tldr_dead` | Find unreachable/dead code |
| `tldr_imports` | Parse imports from a file |
| `tldr_importers` | Find all files importing a module |
| `tldr_change_impact` | Find tests affected by changes |
| `tldr_diagnostics` | Type-check and lint diagnostics |
## Usage
After setup, restart Claude Code. The tools will be available automatically. Example prompts:
- "Use tldr_tree to show the structure of this project"
- "Use tldr_context to explain the `handleAuth` function"
- "Use tldr_impact to see what calls `processPayment`"
TDQS
Scored across 18 tools
Each tool has a clearly distinct and well-defined purpose, from architecture detection to semantic search. There is no overlap or ambiguity; even related tools like tldr_calls and tldr_impact are differentiated by directionality.
All tools follow a consistent pattern: the prefix 'tldr_' followed by a snake_case name (e.g., tldr_change_impact). The naming convention is uniform and predictable.
With 18 tools, the server covers a broad range of static analysis features without being overwhelming. Each tool serves a specific purpose, and the count is well-suited for a comprehensive code understanding toolset.
The tool set is remarkably thorough, covering file tree, structure, imports, call graphs, control and data flow, diagnostics, search, dead code, and more. There are no obvious gaps for the domain of code analysis and comprehension.