reforge-mcp

An MCP server that connects Claude Code to your codebase and systematically cleans it up — scanning, planning, and applying fixes one reversible file at a time.

What it does

Vibe-coded repos accumulate fast: functions nobody calls, copy-pasted logic in three files, one 900-line god module that does everything, circular imports that break at runtime. Spotting this manually takes hours; fixing it safely takes longer. reforge-mcp exposes your codebase to Claude Code as a set of structured tools so it can analyse, prioritise, and apply fixes autonomously — one atomic commit at a time.

The 10 MCP tools

Status

The five analyses scan_repo reports — dead code, duplicates, god files, the dependency graph and cycle detection — are each covered by tests and verified by running the scanner against this repository. Python is the best-exercised language by a distance; the JavaScript, TypeScript and Go adapters share the same machinery but have far less test coverage, so treat their findings as indicative. write_fix will not apply a diff that does not match the file.

Also included

• Health score trending — every get_health_score call appends to health_history so you can chart improvement over time
• Persistent memory — reforge-state.json survives across Claude Code sessions; stores the architecture hypothesis, fix log and pending work
• Monorepo support — scan_repo detects Python, JavaScript, TypeScript, and Go subprojects from their marker files and returns the correct test_command per subproject
• Changelog generation — REFORGE_CHANGES.md is written on demand with score deltas, applied fixes, and pending items

Installation

Prerequisites: Python 3.11+, Claude Code CLI, Git

From PyPI (recommended)

pip install reforge-mcp

From source

git clone https://github.com/codewithaiyan/reforge-mcp.git
cd reforge-mcp
python3 -m venv venv
source venv/bin/activate          # Windows: venv\Scripts\activate
pip install -e .

Register it with Claude Code

claude mcp add reforge -- reforge-mcp

If reforge-mcp is not on your PATH — which is usual when you installed into a virtualenv — give the absolute path instead:

which reforge-mcp                 # Windows: where reforge-mcp
claude mcp add reforge -- /path/from/the/line/above

claude mcp list should then show reforge, and the ten tools become available in your next session. No API key is needed: the server does no network calls and Claude Code supplies the model.

{
  "mcpServers": {
    "reforge": {
      "command": "/absolute/path/to/reforge-mcp",
      "args": [],
      "env": {}
    }
  }
}

Usage

Example session with Claude Code

You: Use reforge-mcp to scan my repo at /home/me/myproject and fix the top issues.

Claude: I'll start with a full scan.

[calls scan_repo("/home/me/myproject")]

The scan found:
- 14 dead functions across 6 files
- 3 duplicate implementations of `calculate_discount`
- 1 god file: src/api/handlers.py (847 lines, 23 exports)
- 2 circular dependencies: auth → middleware → auth

Health score: 47/100

Let me fetch the duplicate implementations to confirm they're identical.

[calls get_chunk("src/billing/utils.py", start_line=12, end_line=28)]
[calls get_chunk("src/cart/helpers.py", start_line=5, end_line=21)]

Confirmed — identical bodies. I'll keep the one in src/billing/utils.py and have
src/cart/helpers.py import it.

[calls write_fix("src/cart/helpers.py", diff_str="...", test_command="pytest")]
[calls git_commit("/home/me/myproject", files=["src/cart/helpers.py"], message="refactor: drop duplicate calculate_discount")]

Committed: a3f91c2. Running next fix...

Sample scan_repo output

Real output, not an illustration. The repository behind it is four files: calculate_discount copy-pasted into src/pricing.py and src/cart.py, an unused parse_legacy_csv and _old_validate, and auth.py and middleware.py importing each other.

{
  "summary": {
    "total_files": 4,
    "total_functions": 7,
    "total_classes": 0,
    "total_imports": 3,
    "total_lines": 41,
    "languages_detected": ["python"]
  },
  "dead_code": [
    { "symbol": "_old_validate",    "file": "auth.py",        "line": 8 },
    { "symbol": "checkout",         "file": "src/cart.py",    "line": 10 },
    { "symbol": "parse_legacy_csv", "file": "src/pricing.py", "line": 7 }
  ],
  "duplicates": [
    {
      "hash": "0a764010605b675494ce6d70c89bb5da7bf02f63589dd69d232c1b057c76569f",
      "locations": [
        { "file": "src/cart.py",    "line": 4, "name": "calculate_discount" },
        { "file": "src/pricing.py", "line": 1, "name": "calculate_discount" }
      ]
    }
  ],
  "god_files": [],
  "dep_graph": {
    "nodes": ["auth.py", "middleware.py", "src/cart.py", "src/pricing.py"],
    "edges": [
      { "from": "auth.py", "to": "middleware.py" },
      { "from": "middleware.py", "to": "auth.py" }
    ],
    "circular": [["auth.py", "middleware.py", "auth.py"]]
  },
  "subprojects": []
}

checkout is reported dead because nothing in those four files calls it — which is what the tool promises, and a reminder that a finding is a question for Claude to answer, not an instruction. On a monorepo, subprojects fills in from the marker files it finds:

"subprojects": [
  { "root": "frontend",        "language": "javascript", "test_command": "npm test" },
  { "root": "services/worker", "language": "go",         "test_command": "go test ./..." }
]

Paths use your platform's separator, so they come back with backslashes on Windows.

Sample REFORGE_CHANGES.md

# Reforge Changes

*Generated: 2025-05-02T14:33:00+00:00*

**Health score:** 73.5 (+26.5 from previous scan)

## Fixes Applied

- refactor: consolidate duplicate calculate_discount
- fix: remove dead parse_legacy_csv from src/importers/csv.py
- refactor: split handlers.py into auth_handlers.py and order_handlers.py

## Pending Fixes

- resolve circular dep: src/auth/__init__.py ↔ src/middleware/auth.py
- remove _old_validate from src/models/user.py

How it works

Reforge works in three phases:

1. Scan

scan_repo walks the repository with parse_directory, feeding each source file to the appropriate tree-sitter adapter (Python, JavaScript/TypeScript, or Go). Tree-sitter builds a full AST — unlike regex, it understands nested scope, decorators, arrow functions, and generics. The adapters extract functions, classes, and imports into typed dataclasses.

Each parsed file also reports every name it uses, collected from the AST while skipping each definition's own name and the contents of comments and strings. From the parse results, three analyses run:

• Dead code — functions whose name appears in nobody's reference set. Matching is by name, not resolved symbol, so a repo-wide name collision hides a finding: a missed one costs nothing, a wrong one costs you working code. Exempt: dunder methods, conventional entry points, test_*, anything listed in __all__, and anything carrying a decorator that is not a pure modifier like @property — a framework calls its registered functions without naming them anywhere a parser can see
• Duplicates — functions whose normalised bodies share the same SHA-256 hash. One-line methods that several classes implement identically are treated as interface conformance, not copy-paste
• God files — non-test files over 500 lines or with more than 10 top-level definitions. Methods do not count towards that total

A dependency graph is built from resolved imports — from x import y, bare import x, relative from ..pkg import y, and JS/TS ./path forms — and checked for cycles with DFS. Imports that resolve outside the repository are external and get no edge. Monorepo subprojects are discovered by scanning subdirectories for marker files (pyproject.toml, package.json, go.mod, tsconfig.json).

2. Plan

Claude Code receives the structured scan report and uses get_chunk to read specific line ranges before deciding what to fix and in what order. The architecture hypothesis, pending fix queue, and health history are all persisted in reforge-state.json via read_memory / write_memory so the plan survives a session restart.

3. Fix

write_fix applies a unified diff atomically:

1. Creates a .bak backup of the target file
2. Parses the diff into hunks and checks every hunk's context and removed lines against the file at the position its header claims
3. Applies the hunks with line-offset tracking — or, if any hunk failed that check, writes nothing at all
4. Runs the optional test command (validated against the allowlist)
5. On any failure — a diff that does not match the file, test exit ≠ 0 — restores from .bak and returns rolled_back: true

git_commit stages only the listed files and creates an atomic commit via GitPython. It reads session_budget and confirm_every from reforge.toml and enforces them: once the session budget is exceeded commits are blocked; at every confirm_every checkpoint the tool returns needs_confirmation: true and waits for confirmed: true before proceeding. The count is held per repository for the life of the server process, so restarting the server starts a new session.

Security

Contributing

# Clone and install
git clone https://github.com/codewithaiyan/reforge-mcp.git
cd reforge-mcp
python3 -m venv venv && source venv/bin/activate
pip install -e ".[dev]"

# Run the full test suite
pytest

# Run with coverage
pytest --cov=src --cov-report=term-missing

Project structure

reforge-mcp/
├── src/reforge_mcp/
│   ├── server.py              # FastMCP server — registers all 10 tools
│   ├── tools/
│   │   ├── scan.py            # scan_repo, get_health_score, monorepo detection
│   │   ├── chunk.py           # get_chunk — line-range file reads
│   │   ├── fix.py             # write_fix — diff apply + rollback
│   │   └── git.py             # git_commit — atomic commits with budget enforcement
│   ├── scanner/
│   │   ├── parser.py          # tree-sitter adapters (Python, JS/TS, Go) + reference collection
│   │   ├── dead_code.py       # unreferenced symbol detection
│   │   ├── duplicates.py      # body-hash duplicate detection
│   │   ├── architecture.py    # project-shape inference
│   │   ├── risk.py            # per-symbol change risk
│   │   └── adapters/          # per-language project-detection stubs
│   └── utils/
│       ├── state.py           # reforge-state.json load/save, gitignore, changelog
│       ├── security.py        # path validation, file guard, command allowlist
│       └── diff.py            # unified diff parser and applier
├── tests/                     # one file per module, plus test_end_to_end.py
│   ├── test_parser.py         #   which drives scan -> fix -> commit for real
│   ├── test_dead_code.py
│   ├── test_duplicates.py
│   ├── test_dep_graph.py
│   ├── test_god_files.py
│   ├── test_diff.py
│   ├── test_fix.py
│   ├── test_git.py
│   ├── test_chunk.py
│   ├── test_memory.py
│   ├── test_monorepo.py
│   ├── test_scanner.py
│   └── test_security.py
├── pyproject.toml
└── reforge.toml               # optional per-repo config

reforge.toml reference

[scan]
ignore = ["generated", "vendor"]   # added to the built-in skip list

[fix]
confirm_every  = 5       # pause for confirmation every N commits
session_budget = 20      # max commits per session, per repo, before a hard stop

Those three keys are the whole file. Every one of them is read.

Changelog

See CHANGELOG.md. If you used 0.1.0, re-scan: three of the four analyses returned meaningless output.

License

MIT — see LICENSE.