AdoReviewLens

Utilities for fetching active Azure DevOps PR comments via MCP interface.

Requirements

• Python 3.10+
• Azure DevOps Personal Access Token with Code (read) and Pull Requests (read)

Installation

Local development

python3 -m venv .venv
source .venv/bin/activate
pip install -e .

cp .env.example .env
# edit .env with your AZDO_ORG_URL, AZDO_PAT, AZDO_PROJECT, AZDO_REPO

Alternatively, if you prefer uv:

uv venv
source .venv/bin/activate
uv pip install -e .

Running python -m ado_review_lens.cli --help will confirm the CLI is available.

CLI usage

source .venv/bin/activate
python -m ado_review_lens.cli --pr 123
python -m ado_review_lens.cli --url "https://dev.azure.com/org/project/_git/repo/pullrequest/123"
python -m ado_review_lens.cli --pr 123 --allow-cross-project

HTTP API server

source .venv/bin/activate
uvicorn ado_review_lens.api:app --reload
# POST /api/v1/pr/comments with JSON {"prId": 123}

MCP server

source .venv/bin/activate
python -m ado_review_lens.server

Use the resulting stdio endpoint with the Model Context Protocol client of your choice (e.g., Claude Desktop or MCP Inspector).

Installing via mcp

The official MCP tooling expects uv to be available:

uv tool install mcp

# Register the server with an MCP client (e.g., Claude Desktop)
uv run mcp install src/ado_review_lens/server.py --name "ADO Review Lens"

# Launch the MCP Inspector for interactive testing
uv run mcp dev src/ado_review_lens/server.py --with-editable .

The install command accepts -v KEY=value or -f .env flags if you want to supply Azure DevOps environment variables during registration.

Manual MCP client registration

Some MCP clients accept explicit stdio configurations:

{
  "mcpServers": {
    "ado-review-lens": {
      "command": "python", // or "python3" or path to python interpreter
      "args": [
        "-m",
        "ado_review_lens.server"
      ],
      "env": {
        "AZDO_ORG_URL": "https://dev.azure.com/<org>",
        "AZDO_PAT": "<personal_access_token>",
        "AZDO_PROJECT": "<default_project>",
        "AZDO_REPO": "<default_repo>"
      }
    }
  }
}

Merge this snippet with existing definitions as needed, then restart or refresh your client so it discovers the server.

Other MCP clients

The repository ships with a server.json definition describing the PyPI package and stdio transport. Any MCP-compatible client that accepts server.json entries can load the server by either:

1. Copying server.json into the client’s configuration directory, or
2. Merging the io.github.om-surushe/ado-review-lens entry into an existing MCP registry file.

The critical values to replicate are:

{
  "name": "io.github.om-surushe/ado-review-lens",
  "transport": {
    "type": "stdio",
    "command": ["python", "-m", "ado_review_lens.server"]
  }
}

Ensure the environment variables required by Azure DevOps (PAT, org URL, project, repo) are set in the client before launching the server process.