@smartcase/mcp

SmartCase MCP server — expose AI-generated QA test suites to Cursor, Claude Desktop, and CI/CD pipelines via the Model Context Protocol.

Quick Start

1. Create an API Key

Go to Settings → API Keys in your SmartCase account and create a new key (sk-smrt-...).

2. Configure your IDE

Cursor / Claude Desktop

Add to your mcp.json (Cursor: ~/.cursor/mcp.json, Claude Desktop: ~/Library/Application Support/Claude/claude_desktop_config.json):

{
  "mcpServers": {
    "smartcase": {
      "command": "npx",
      "args": ["-y", "@smartcase/mcp"],
      "env": {
        "FLOWSPECQA_API_KEY": "sk-smrt-your-key-here",
        "FLOWSPECQA_BASE_URL": "https://www.flowspecqa.com"
      }
    }
  }
}

Restart Cursor/Claude and the SmartCase tools will appear automatically.

Available Tools

| Tool | What it does | |---|---| | list_entities | List entities (HU-xxx) in a project with their spec review status | | get_test_suite | Fetch all generated test files for a run ID | | generate_suite | Trigger AI generation for an entity, returns a run_id | | setup_and_run | Download suite + generate playwright.config.ts with HTML & JSON reporters | | push_results | Push test execution results back to SmartCase |

Example agent prompt (Cursor)

Use SmartCase to generate and run a Playwright suite for entity HU-042 in project abc-123
against https://staging.myapp.com. Push the results when done.

The agent will: generate_suite → setup_and_run → write files → run Playwright → push_results.

CI/CD — GitHub Actions

name: SmartCase + Playwright

on: [push]

jobs:
  run-tests:
    runs-on: ubuntu-latest
    steps:
      - uses: actions/checkout@v4

      - name: Fetch test suite from SmartCase
        env:
          FLOWSPECQA_API_KEY: ${{ secrets.FLOWSPECQA_API_KEY }}
          FLOWSPECQA_BASE_URL: https://www.flowspecqa.com
        run: |
          npx -y @smartcase/mcp get-suite \
            --run-id ${{ vars.SMARTCASE_RUN_ID }} \
            --output ./tests/generated

      - name: Install Playwright
        run: npx playwright install --with-deps

      - name: Run tests
        run: npx playwright test ./tests/generated --reporter=json > results.json

      - name: Push results to SmartCase
        env:
          FLOWSPECQA_API_KEY: ${{ secrets.FLOWSPECQA_API_KEY }}
        run: |
          npx @smartcase/mcp push-results \
            --run-id ${{ vars.SMARTCASE_RUN_ID }} \
            --results-file results.json

Environment Variables

| Variable | Required | Default | |---|---|---| | FLOWSPECQA_API_KEY | ✅ Yes | — | | FLOWSPECQA_BASE_URL | No | https://www.flowspecqa.com |

Local Development

git clone https://github.com/pablomra/smartcase.git
cd flowspecqa-mcp
npm install
npm run build

# Test with MCP Inspector
FLOWSPECQA_API_KEY=sk-smrt-... npx @modelcontextprotocol/inspector node dist/index.js

Score Semantics

| Score | Meaning | |---|---| | Design Score | Code quality · maintainability (0–100) | | Run-Ready Score | 100 = zero BLOCKERs, safe to execute; <100 = fix issues first |

get_test_suite will warn you if the suite has BLOCKERs before you run it.