AiderSAFE
Overview
From the repository's own README, as read at the audited commit. Badges and raw HTML are left out.
A Model Context Protocol (MCP) server that allows Claude and other MCP clients to connect to Aider for efficient file editing capabilities.
Overview
This MCP server bridges the gap between AI assistants like Claude and Aider's powerful file editing capabilities. It provides a standardized interface through the Model Context Protocol, allowing Claude to:
- Edit existing files using Aider's capabilities
- Create new files with content
- Extract code blocks from markdown text
- Get git status information
- Check Aider installation status
- Access and use Aider's configuration system
Prerequisites
- Python 3.8 or higher
- Aider installed (
pip install aider-chat) - An API key for OpenAI or Anthropic (depending on which model you want Aider to use)
- Git repository for file editing
Installation
You can install the Aider MCP server in several ways:
Using UV (Recommended)
If you have uv installed:
# Install uv if you don't have it curl -fsSL https://astral.sh/uv/install.sh | bash # Run directly with uvx (no installation required) uvx aider-mcp
Using PIP
# Install the package pip install aider-mcp # Run the server aider-mcp
Usage
The Aider MCP server runs in MCP protocol mode over stdio by default, which is designed for direct integration with MCP clients like Claude Desktop and Cursor IDE.
# Run directly with uvx (recommended) uvx aider-mcp # With repository path specified uvx aider-mcp --repo-path=/path/to/your/repo # With custom Aider executable path uvx aider-mcp --aider-path=/path/to/aider # With environment variables REPO_PATH=/path/to/your/repo uvx aider-mcp
Command Line Options
You can customize the server with these environment variables or command-line arguments:
--aider-path: Path to the Aider executable (default: "aider", automatically searches PATH)--repo-path: Path to the g
71e3ca583e59OBSERVED · 2026-10-08Connect
Built from this server's own package name, version and transport as found in its source — not copied from anyone's documentation, so it cannot drift against a page we do not control.
claude mcp add aider-mcp -- uvx aider-mcp
{
"mcpServers": {
"aider-mcp": {
"command": "uvx",
"args": [
"aider-mcp"
]
}
}
}Exposed tools (6)
4 read · 2 write · 0 destructive.
| Tool | Risk | Description |
|---|---|---|
aider_config | read | Get detailed Aider configuration information. Use this to:\n\n |
aider_status | read | Check the status of Aider and its environment. Use this to:\n\n |
create_files | write | Create new files in a git repository. Use this when you need to:\n\n |
edit_files | write | AI pair programming tool for making targeted code changes. Use this tool to:\n\n |
extract_code | read | Extract code blocks from markdown or text. Use this to:\n\n |
git_status | read | Get the current git status of a repository. Shows modified, untracked, and staged files.\n\n |
Trust audit
SAFEgrade B · trust 89/100 Nothing in the source contradicts what it says it does. Grade A is reserved for packages that have also passed the behavioural sandbox.
| Layer | What it checks | Result |
|---|---|---|
| L0 | Provenance & inventory | PASS |
| L1 | Static analysis of the code | PASS |
| L2 | Instruction surface (what it tells the agent) | PASS |
| L3 | Class-specific surface | PASS |
| L4 | Behavioural (sandbox) | SKIPPED |
What the source does
- Filesystem
- declared (1 observation(s))
- Network
- none-observed
- Shell
- none-observed
- Dependencies
- pinned
- Secrets in source
- none-found
Findings (1)
curl -fsSL https://astral.sh/uv/install.sh | bash
Gates applied: no_behavioural_pass.
71e3ca583e59full audit observations/trust-audit/mcp-server/sengokudaikon__aider.json · Report an issue / request a re-scanAudit history
Every audit this server has had. A grade with a past is a grade somebody is still checking.
| Date | Source | Verdict | Grade | Score | Change |
|---|---|---|---|---|---|
| 2026-10-08 | 71e3ca583e59 | SAFE | B | 89 | first audit |
Questions
What tools does Aider expose?
6 in total: 4 read-only, 2 that write, and 0 that can delete or overwrite. Every one is listed on this page with its risk.
Is Aider safe to connect to an agent?
The audit found nothing in the source that contradicts what it says it does, and graded it B (89/100). Grade A is held back for packages that have also passed a sandboxed behavioural run, which is why a clean server reads B.
What credentials does Aider need?
No credential environment variables were found in its source, so it appears to need none.
How does Aider run?
It speaks stdio, so it runs as a local process your client starts. It is published on PyPI as aider-mcp.
How current is this page?
The grade is for one exact copy of the source (71e3ca583e59), read on 2026-10-08. The repository is watched and re-audited when it changes.