Lets AI clients run git status, log, diff, branch, and show commands and receive typed, schema-validated JSON instead of raw terminal text.
Also known as Dave-London/Pare, @paretools/git
- Stars
- 141
- Tools
- 15
- Downloads, 30 days
- —
- Latest version
- 0.8.0
- Last commit
- Nov 10, 2026
- License
- MIT
What is Pare Git MCP?
Pare Git lets AI clients run git status, log, diff, branch, and show commands and receive typed, schema-validated JSON instead of raw terminal text. It exposes 15 tools, such as status, log, diff.
Summary generated from the project's README
About Pare Git MCP server
Providing structured, schema-validated JSON for version control operations is what this server enables for AI agents. Instead of forcing an AI to parse fragile, platform-dependent terminal text, it converts standard commands like git status, log, diff, branch, and show into typed data. This approach prevents errors caused by ANSI escape codes or locale-specific formatting that often break regex-based parsing in automated workflows.
Developers use this server locally by connecting it to MCP clients such as Claude Desktop, Cursor, or Windsurf. It is configured via npx and can be customized using environment variables like PARE_GIT_TOOLS to restrict which specific commands are exposed to the agent. By returning structuredContent alongside human-readable text, the server significantly reduces token usage, often by over 50% for commands like git status, making agent interactions more efficient and cost-effective.
What you can do with it
- 1Analyze repository statusAn AI agent can examine staged, modified, and untracked files using a structured JSON object rather than reading raw terminal output.
- 2Reduce token consumptionDevelopers can minimize the cost and context window usage of AI calls by replacing verbose command outputs with compact JSON.
- 3Ensure cross-platform reliabilityAn AI client can perform git operations on different operating systems without failing due to variations in terminal formatting.
Written from the project's README. Every use case is checked against a quote from it.
How to install Pare Git MCP
Formats verified 10 Nov 2026Add to CursorAdd to GooseInstall Pare Git MCP in Claude Code
$ claude mcp add git -- npx -y @paretools/gitInstall Pare Git MCP in Cursor
~/.cursor/mcp.json
{
"mcpServers": {
"git": {
"command": "npx",
"args": [
"-y",
"@paretools/git"
]
}
}
}Install Pare Git MCP in VS Code
.vscode/mcp.json
{
"servers": {
"git": {
"type": "stdio",
"command": "npx",
"args": [
"-y",
"@paretools/git"
]
}
}
}Install Pare Git MCP in VS Code CLI
$ code --add-mcp '{"name":"git","type":"stdio","command":"npx","args":["-y","@paretools/git"]}'Install Pare Git MCP in Claude Desktop
claude_desktop_config.json
{
"mcpServers": {
"git": {
"command": "npx",
"args": [
"-y",
"@paretools/git"
]
}
}
}Install Pare Git MCP in Codex
$ codex mcp add git -- npx -y @paretools/gitInstall Pare Git MCP in Codex config.toml
~/.codex/config.toml
[mcp_servers.git] command = "npx" args = ["-y", "@paretools/git"]
Install Pare Git MCP in Gemini CLI
$ gemini mcp add -s user git npx -- -y @paretools/gitInstall Pare Git MCP in Gemini settings.json
~/.gemini/settings.json
{
"mcpServers": {
"git": {
"command": "npx",
"args": [
"-y",
"@paretools/git"
]
}
}
}Started by HostedMCP on 16 Nov 2026
We ran this command for @paretools/git 0.8.0 in an isolated container with no network and no keys. It answered and listed 15 tools.
Pare Git MCP tools 15
Read from the server with tools/liststatusReturns the working tree status as structured data (branch, staged, modified, untracked, conflicts). Use instead of running `git status` in the terminal.
logReturns commit history as structured data. Use instead of running `git log` in the terminal.
diffReturns file-level diff statistics as structured data. Use full=true for patch content. Use instead of running `git diff` in the terminal.
branchLists, creates, or deletes branches. Returns structured branch data. Use instead of running `git branch` in the terminal.
showShows commit details and diff statistics for a given ref. Use instead of running `git show` in the terminal.
addStages files for commit. Returns structured data with count and list of staged files. Use instead of running `git add` in the terminal.
commitCreates a commit with the given message. Returns structured data with hash, message, and change statistics. Use instead of running `git commit` in the terminal.
pushPushes commits to a remote repository. Returns structured data with success status, remote, branch, and summary. Use instead of running `git push` in the terminal.
pullPulls changes from a remote repository. Returns structured data with success status, summary, change statistics, and any conflicts. Use instead of running `git pull` in the terminal.
checkoutSwitches branches or restores files. Returns structured data with ref, previous ref, and whether a new branch was created. Use instead of running `git checkout` in the terminal.
tagLists tags sorted by creation date. Returns structured tag data with name, date, and message. Use instead of running `git tag` in the terminal.
stash-listLists all stash entries with index, message, and date. Returns structured stash data. Use instead of running `git stash list` in the terminal.
| status | Returns the working tree status as structured data (branch, staged, modified, untracked, conflicts). Use instead of running `git status` in the terminal. | |
| log | Returns commit history as structured data. Use instead of running `git log` in the terminal. | |
| diff | Returns file-level diff statistics as structured data. Use full=true for patch content. Use instead of running `git diff` in the terminal. | |
| branch | Lists, creates, or deletes branches. Returns structured branch data. Use instead of running `git branch` in the terminal. | |
| show | Shows commit details and diff statistics for a given ref. Use instead of running `git show` in the terminal. | |
| add | Stages files for commit. Returns structured data with count and list of staged files. Use instead of running `git add` in the terminal. | |
| commit | Creates a commit with the given message. Returns structured data with hash, message, and change statistics. Use instead of running `git commit` in the terminal. | |
| push | Pushes commits to a remote repository. Returns structured data with success status, remote, branch, and summary. Use instead of running `git push` in the terminal. | |
| pull | Pulls changes from a remote repository. Returns structured data with success status, summary, change statistics, and any conflicts. Use instead of running `git pull` in the terminal. | |
| checkout | Switches branches or restores files. Returns structured data with ref, previous ref, and whether a new branch was created. Use instead of running `git checkout` in the terminal. | |
| tag | Lists tags sorted by creation date. Returns structured tag data with name, date, and message. Use instead of running `git tag` in the terminal. | |
| stash-list | Lists all stash entries with index, message, and date. Returns structured stash data. Use instead of running `git stash list` in the terminal. |
Show all 15 toolsShow fewer tools
stashPushes, pops, applies, or drops stash entries. Returns structured result with action, success, and message. Use instead of running `git stash` in the terminal.
remoteLists remote repositories with fetch and push URLs. Returns structured remote data. Use instead of running `git remote -v` in the terminal.
blameShows commit annotations for a file, grouped by commit. Returns structured blame data with deduplicated commit metadata (hash, author, date) and their attributed lines. Use instead of running `git blame` in the terminal.
| stash | Pushes, pops, applies, or drops stash entries. Returns structured result with action, success, and message. Use instead of running `git stash` in the terminal. | |
| remote | Lists remote repositories with fetch and push URLs. Returns structured remote data. Use instead of running `git remote -v` in the terminal. | |
| blame | Shows commit annotations for a file, grouped by commit. Returns structured blame data with deduplicated commit metadata (hash, author, date) and their attributed lines. Use instead of running `git blame` in the terminal. |
Configuration
PARE_GIT_TOOLSPARE_NPM_TOOLSPARE_TOOLS
| Name | Description | |
|---|---|---|
| PARE_GIT_TOOLS | — | |
| PARE_NPM_TOOLS | — | |
| PARE_TOOLS | — |
Automated checks
Ran Nov 16, 2026Full report →Passed our automated checks. That is not a security audit: read the code and give the server the narrowest access it needs. How checks work
- Source repository linked and reachabledave-london/pare is public and reachablegithub.com/dave-london/pare ↗Pass
- Package names this repository as its source@paretools/git names dave-london/pare as sourcenpmjs.com/package/@paretools/git ↗Pass
- Build provenance (Sigstore attestation)Sigstore attestation for @paretools/gitnpmjs.com/package/@paretools/git/v/0.8.0#provenance ↗Pass
- Open source license (OSI)MITspdx.org/licenses/MIT.html ↗Pass
- Commit or release in the last yearLast commit or release 2026-11-10github.com/dave-london/pare ↗Pass
- OpenSSF ScorecardNot checked
- Package name not close to a popular packageNo popular package one edit away from @paretools/gitPass
- No critical advisory in the package or its dependencies100 resolved dependencies checked; no critical advisoryosv.dev ↗Pass
- No hidden instructions to the model15 tool descriptions and the README read; no override phrases, credential paths or hidden charactersPass
- Destructive tools are labelledNo tool named like delete, drop or execPass
- Remote endpoint answers the MCP handshakeNot checked
README
View on GitHubReliable, structured CLI output for AI agents — no more parsing fragile terminal text.
Pare provides MCP servers that wrap common developer tools (git, npm, docker, test runners, etc.) and return clean, schema-validated JSON instead of raw terminal text. Agents get typed data they can act on directly, without brittle string parsing.
The Problem
Parsing CLI output is fragile. Raw terminal text includes ANSI escape codes, decorative headers, progress bars, locale-specific formatting, and platform differences that break agent workflows in subtle ways. An agent that works fine with git status on macOS may fail on Windows because the output format changed. A test runner's summary line might shift between versions, silently breaking a regex.
Pare eliminates this entire class of errors by returning schema-validated JSON with consistent field names, regardless of platform, tool version, or locale. As a bonus, structured output is significantly smaller — agents use fewer tokens per tool call:
| Tool Command | Raw Tokens | Pare Tokens | Reduction |
|---|---|---|---|
docker build (multi-stage, 11 steps) |
373 | 20 | 95% |
git log --stat (5 commits, verbose) |
4,992 | 382 | 92% |
npm install (487 packages, warnings) |
241 | 41 | 83% |
vitest run (28 tests, all pass) |
196 | 39 | 80% |
cargo build (2 errors, help text) |
436 | 138 | 68% |
pip install (9 packages, progress bars) |
288 | 101 | 65% |
cargo test (12 tests, 2 failures) |
351 | 190 | 46% |
npm audit (4 vulnerabilities) |
287 | 185 | 36% |
Token estimates use ~4 chars/token. The biggest savings appear on verbose commands (builds, installs, tests). For simpler tools like
eslintortsc, the main advantage is reliable structured data — agents can use typed JSON directly rather than parsing strings.
How It Works
Each Pare tool returns two outputs:
content— human-readable text, for MCP clients that display itstructuredContent— typed, schema-validated JSON, ready for agents to process
This uses MCP's structuredContent and outputSchema features to provide type-safe, validated data that agents can rely on without custom parsing.
Example: git status
Raw git output (~118 tokens):
On branch main
Your branch is ahead of 'origin/main' by 2 commits.
(use "git push" to publish your local commits)
Changes to be committed:
(use "git restore --staged <file>..." to unstage)
modified: src/index.ts
new file: src/utils.ts
Changes not staged for commit:
(use "git add <file>..." to update what will be committed)
(use "git restore <file>..." to discard changes in working directory)
modified: README.md
Untracked files:
(use "git add <file>..." to include in what will be committed)
temp.log
Pare structured output (~59 tokens):
{
"branch": "main",
"upstream": "origin/main",
"ahead": 2,
"staged": [
{ "file": "src/index.ts", "status": "modified" },
{ "file": "src/utils.ts", "status": "added" }
],
"modified": ["README.md"],
"deleted": [],
"untracked": ["temp.log"],
"conflicts": [],
"clean": false
}
50% fewer tokens. Zero information lost. Fully typed. Savings scale with output verbosity — test runners and build logs see 80–92% reduction.
Available Servers (28 packages, 240 tools)
Install only the servers relevant to your stack — most projects need just 2–4. The full catalog covers a wide range of ecosystems so Pare works wherever you do.
| Category | Servers | Tools | Wraps |
|---|---|---|---|
| Version Control | git, github | 55 | git, gh |
| Languages & Packages | npm, python, cargo, go, deno, bun, nix, dotnet, ruby, swift, jvm | 101 | npm, pip, cargo, go, deno, bun, nix, dotnet, gem, swift, gradle, maven |
| Build, Lint & Test | build, lint, test, cmake, bazel | 23 | tsc, esbuild, vite, webpack, eslint, prettier, biome, vitest, pytest, jest |
| Infrastructure | docker, k8s, infra, security, remote | 40 | docker, kubectl, helm, terraform, ansible, trivy, ssh |
| Utilities | search, http, make, process, db | 21 | ripgrep, fd, curl, make, just, psql, mysql, redis, mongosh |
Tool Schemas — detailed response examples and field descriptions for every tool. See also Tool Response Examples for quick JSON samples.
Quick Setup
# 1. Configure MCP servers (non-interactive)
npx @paretools/init --client claude-code --preset web
# 2. Add agent rules to your project
# (append to existing CLAUDE.md, or copy if new)
cat node_modules/@paretools/init/rules/CLAUDE.md >> CLAUDE.md
# 3. Restart your client session
# 4. Validate
npx @paretools/init doctor
Available presets: web, python, rust, go, jvm, dotnet, ruby, swift, mobile, devops, full
Setup Guides by Client
| Claude Code | Claude Desktop | Cursor |
| VS Code / Copilot | Windsurf | Cline / Roo Code |
| OpenAI Codex | Gemini CLI | Zed |
| Continue.dev |
Full Quickstart Guide — presets, ecosystem mapping, validation
Manual Configuration — config paths and formats for all clients
Agent Integration Guide — rule files, hooks, CLI-to-MCP mapping
Configuration
Tool Selection
By default, every Pare server registers all of its tools. If a server exposes tools you don't need — or you want to limit which tools are available to an agent — you can filter them with environment variables.
Per-server filter — restrict a single server's tools:
# Only register status and log in the git server
PARE_GIT_TOOLS=status,log npx @paretools/git
Universal filter — restrict tools across all servers:
# Only register these specific tools across any server
PARE_TOOLS=git:status,git:log,npm:install npx @paretools/git
Disable all tools — set the env var to an empty string:
PARE_GIT_TOOLS= npx @paretools/git # no tools registered
| Env Var | Scope | Format | Example |
|---|---|---|---|
PARE_TOOLS |
All servers | server:tool,... |
git:status,npm:install |
PARE_{SERVER}_TOOLS |
One server | tool,... |
status,log,diff |
Rules:
- No env var = all tools enabled (default)
PARE_TOOLS(universal) takes precedence over per-server vars- Server names use uppercase with hyphens replaced by underscores (e.g.,
PARE_MY_SERVER_TOOLS) - Whitespace around commas is ignored
Common patterns:
# Read-only git (no push, commit, add, checkout)
PARE_GIT_TOOLS=status,log,diff,branch,show
# Minimal npm
PARE_NPM_TOOLS=install,test,run
# Only specific tools across all servers
PARE_TOOLS=git:status,git:diff,npm:install,test:run
In JSON MCP config, set via the env key:
{
"mcpServers": {
"pare-git": {
"command": "npx",
"args": ["-y", "@paretools/git"],
"env": {
"PARE_GIT_TOOLS": "status,log,diff,show"
}
}
}
}
Troubleshooting
| Issue | Solution |
|---|---|
npx not found / ENOENT on Windows |
Use cmd /c npx wrapper (see your client's setup guide) |
| Slow first start | Run npx -y @paretools/git once to cache, or install globally: npm i -g @paretools/git |
| Node.js version error | Pare requires Node.js >= 20 |
| NVM/fnm PATH issues | Use absolute path to npx: e.g., ~/.nvm/versions/node/v22/bin/npx |
| MCP connection timeout | Set MCP_TIMEOUT=30000 for Claude Code, or increase initTimeout in client config |
| Too many tools filling context | Use tool selection env vars to limit tools, or only install the servers you need |
Contributing
Each server is a self-contained package. See CONTRIBUTING.md for the full guide.
License
Pare Git MCP questions
How do I add Pare Git MCP to Claude Code?
Run claude mcp add git -- npx -y @paretools/git in your terminal. Add --scope user to make it available in every project.
How do I add Pare Git MCP to Cursor?
Open ~/.cursor/mcp.json and add the server under mcpServers, using the Cursor config shown in the install section above.
What tools does Pare Git MCP provide?
It exposes 15 tools, including status, log, diff, branch, show.
What does the git MCP server return?
It returns structured, typed JSON for git operations such as status, log, diff, branch, and show, plus human-readable content for display.
How do I limit which git tools are registered?
Set the PARE_GIT_TOOLS environment variable to a comma-separated list of tool names, for example status,log,diff.
Does the git server use structured output schemas?
Yes, it uses MCP's structuredContent and outputSchema features so agents get validated, type-safe data.
How does the git server reduce token usage?
It converts verbose raw git output into compact JSON, cutting tokens significantly; the README cites roughly 92% fewer tokens for a verbose git log.