Best for
- Developers who need AI assistants to understand local or remote codebases.
- Code-analysis workflows that need control over token counts and file scope.
- Users of Claude Desktop, VS Code, Cline, Cursor, or Claude Code.
R
67FMRS · C
The supplied material identifies Repomix as an MCP server provided by its upstream project. It runs through stdio and offers local and remote repository packaging, output reading and search, and security-checked filesystem reads. Authentication, write access, and private-repository support are not documented.
Repomix is a CLI tool and MCP server that packages local or remote GitHub repositories into consolidated outputs for AI-assisted code analysis. It supports XML, Markdown, JSON, and plain text, plus compression, token counting, file filtering, Git logs, and diffs.
Tools
pack_codebase
Package a local code directory into a consolidated XML file.
attach_packed_output
Attach an existing Repomix packed output file.
pack_remote_repository
Fetch, clone, and package a GitHub repository.
read_repomix_output
Read a Repomix output file, optionally by line range.
grep_repomix_output
Search a Repomix output with JavaScript regular expressions.
file_system_read_file
Read a local file with sensitive-information security checks.
file_system_read_directory
List the contents of a local directory.
Setup
Add the configuration below to an MCP-compatible client, then start the Repomix MCP server with npx -y repomix --mcp. Docker can also be used with ghcr.io/yamadashy/repomix.
claude_desktop_config.json
{"mcpServers":{"repomix":{"command":"npx","args":["-y","repomix","--mcp"]}}}
Fit and risk
Not for
- Users needing database, cloud-storage, or browser-automation capabilities.
- Workflows requiring direct codebase modification or general write operations.
- Users needing authentication, private-repository access, or capabilities not documented in the supplied material.
Required permissions
- Read the specified local directories and files.
- Fetch, clone, and process the specified GitHub repositories.
- Access existing Repomix output files.
- Run Secretlint-sensitive-data checks; remote configuration is not loaded unless explicitly trusted.
Risks and side effects
- Packed content may contain sensitive information; despite Secretlint checks, review outputs before sending them to AI tools.
- Remote repository configuration is untrusted by default; enabling trust may run external processing commands defined in configuration.
- Large repositories may exceed AI-tool file-size or context limits and require filtering, compression, or output splitting.
Troubleshooting
- Confirm the client uses npx -y repomix --mcp, or verify that the Docker command is available.
- Ensure local paths are absolute and that the target directory or output file exists.
- Check .gitignore, .ignore, .repomixignore, and include/ignore patterns for unintended exclusions.
- For remote repositories, verify the repository URL, branch or commit identifier, and remote-config trust settings.
Use cases
Prepare repository context for AI code reviews, refactoring, documentation, and test generation.
Filter, compress, or search large codebases by pattern.
Analyze remote GitHub repositories directly.
Read existing packed outputs and retrieve content incrementally.
Supported clients
Claude DesktopFull support
VS CodeFull support
ClineFull support
CursorFull support
Claude CodeFull support