ProofFlow MCP Server
MCP (Model Context Protocol) server that exposes ProofFlow's audit capabilities as tools for Claude Code and Codex.
What it does
Lets AI coding assistants trigger ProofFlow workflows directly:
- Scan folders — index files with SHA-256 hashes, create audit Cases
- Code review — analyze git diffs, generate evidence-backed risk claims
- Suggest actions — generate file organization suggestions
- Approve & execute — run pending actions (with policy gate awareness)
- Export Proof Packets — generate shareable markdown audit reports
- Search — full-text search across indexed artifacts
- Undo — reverse executed actions
Prerequisites
- Python 3.11+
- ProofFlow backend running on
http://127.0.0.1:8787
Installation
cd mcp-server
pip install -e .
For development (includes test dependencies):
pip install -e ".[dev]"
Note: Install in a separate virtualenv from the backend to avoid starlette version conflicts between FastAPI and the MCP SDK.
Configuration
Claude Code
Add to your project's .mcp.json or ~/.claude/settings.json:
{
"mcpServers": {
"proofflow": {
"command": "python",
"args": ["-m", "proofflow_mcp"],
"env": {
"PROOFFLOW_BASE_URL": "http://127.0.0.1:8787"
}
}
}
}
Claude Desktop
Add to claude_desktop_config.json:
- macOS:
~/Library/Application Support/Claude/claude_desktop_config.json - Windows:
%APPDATA%\Claude\claude_desktop_config.json
{
"mcpServers": {
"proofflow": {
"command": "proofflow-mcp",
"env": {
"PROOFFLOW_BASE_URL": "http://127.0.0.1:8787"
}
}
}
}
Codex
Add to your Codex MCP configuration with the same command/args pattern.
Environment Variables
| Variable | Default | Description |
|---|---|---|
PROOFFLOW_BASE_URL |
http://127.0.0.1:8787 |
ProofFlow backend URL |
PROOFFLOW_TIMEOUT |
30 |
HTTP request timeout in seconds |
PROOFFLOW_API_KEY |
(unset) | API key for backend auth (optional) |
PROOFFLOW_MCP_MAX_CONCURRENT |
5 |
Max concurrent MCP tool calls |
Available Tools
| Tool | Description |
|---|---|
proofflow_health |
Check backend connectivity |
proofflow_scan |
Scan a folder, create Case + Artifacts |
proofflow_suggest |
Generate cleanup suggestions for a scanned Case |
proofflow_review |
AgentGuard code review on a git repo |
proofflow_status |
Get full Case status (artifacts, claims, actions) |
proofflow_approve_execute |
Approve and execute a pending action |
proofflow_export_packet |
Export Case as Proof Packet (markdown) |
proofflow_search |
Full-text search across artifacts |
proofflow_list_cases |
List all Cases |
proofflow_list_actions |
List actions for a Case |
proofflow_undo |
Undo an executed action |
proofflow_decide |
Create a decision to approve/reject a policy gate |
Usage Example
In Claude Code, once configured:
> Review the code changes in this repo for safety
Claude will call proofflow_review with the current repo path,
creating an AgentGuard Case with evidence-backed claims.
Running Tests
cd mcp-server
python -m pytest tests/ -v
Running the Server Directly
python -m proofflow_mcp
The server communicates via stdio (stdin/stdout) per the MCP protocol.
Architecture
Claude Code / Codex
↓ (MCP stdio)
ProofFlow MCP Server
↓ (HTTP)
ProofFlow Backend (localhost:8787)
↓
SQLite DB + Local Files
The MCP server is a thin adapter — all business logic lives in the ProofFlow backend.
Release files for proofflow-mcp 0.1.2
For a detailed explanation of source distributions (sdists) and built distributions (wheels), please see the package formats documentation.
Source distribution (sdist)
| File | Size | Uploaded | |
|---|---|---|---|
| proofflow_mcp-0.1.2.tar.gz | 11.5 kB | Details |
Built distribution (wheel)
| File | Interpreter | ABI | Platform | Reset |
|---|---|---|---|---|
| proofflow_mcp-0.1.2-py3-none-any.whl | Python 3 | none | any | Details |
Total release size: 21.8 kB
Release files / proofflow_mcp-0.1.2.tar.gz
| Download URL | proofflow_mcp-0.1.2.tar.gz |
|---|---|
| Size | 11.5 kB |
| Tags | Source |
|
SHA-256 checksum How to use checksums |
e45ef6742c9c285388adaa188d1ac6920fb3e0026d8eb7cfb0dd5b0a40b76b7f
|
|
BLAKE2b-256 checksum How to use checksums |
aa3cdfd5bb87f28be389bfa126a1f2d433001364704794d75e09ffb93d25bcb5
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
Yes |
| Uploaded via |
twine/6.1.0 CPython/3.13.12
|
Provenance
Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.
PyPI Publish Attestation
PyPI verified that this artifact, at this checksum, originated from the publisher listed below.
Signed by GitHub Actions, verified by PyPI on May 6, 2026.
Transparency logRelease files / proofflow_mcp-0.1.2-py3-none-any.whl
| Download URL | proofflow_mcp-0.1.2-py3-none-any.whl |
|---|---|
| Size | 10.3 kB |
| Tags | Python 3 |
|
SHA-256 checksum How to use checksums |
f1b6b7cf47eef33b17788fe25c2fb55e40173359731f28295a7c7d25501eab9c
|
|
BLAKE2b-256 checksum How to use checksums |
a8d3f93c578b8d76ab6efc13d0a1ec2d3ebf884ed6a537347d0348e9f73e8c0c
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
Yes |
| Uploaded via |
twine/6.1.0 CPython/3.13.12
|
Provenance
Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.
PyPI Publish Attestation
PyPI verified that this artifact, at this checksum, originated from the publisher listed below.
Signed by GitHub Actions, verified by PyPI on May 6, 2026.
Transparency log