Add contribution guide with recommended debugging workflow

CONTRIBUTING.md

@@ -0,0 +1,59 @@
+# Contributing to FlowDex
+
+Thank you for investing time in improving FlowDex! This guide summarizes the expectations for local development and the recommended debugging workflow used by the project.
+
+## Local Development Setup
+
+1. **Create a virtual environment** (Python 3.10+):
+   ```bash
+   python -m venv .venv
+   source .venv/bin/activate
+   ```
+2. **Install the application dependencies**:
+   ```bash
+   pip install -r server/requirements.txt
+   ```
+3. **Install developer tooling** (tests + UI automation):
+   ```bash
+   pip install pytest pytest-asyncio httpx==0.25.2 playwright
+   playwright install  # downloads browser binaries
+   ```
+4. (Optional) **Run the API locally** for manual testing:
+   ```bash
+   uvicorn server.app:app --reload --port 8787
+   ```
+5. (Optional) **Start the full docker-compose stack** when you need Redis + the MCP bridge:
+   ```bash
+   docker compose up --build
+   ```
+
+## Recommended Debugging Workflow
+
+Follow this workflow whenever you reproduce or validate a bug fix. It keeps feedback loops fast and avoids jumping straight to manual reproduction.
+
+1. **Start with automated tests and logs.**
+   - Run the full suite with `pytest server/tests`.
+   - When the docker-compose stack is running, stream API logs with `docker compose logs -f api` and Redis logs with `docker compose logs -f redis` to surface server-side errors quickly.
+   - For CLI regressions, run targeted scripts from the `examples/` directory to capture console output.
+2. **Use Playwright for UI confirmation.**
+   - Add or update Playwright tests alongside the feature you are touching.
+   - Execute them with `pytest tests/ui` (or any other directory you create for UI checks) to confirm behaviour automatically.
+3. **Capture targeted Playwright scripts when results are ambiguous.**
+   - Write a focused script that reproduces the user flow in question.
+   - Save a screenshot or trace artifact (e.g., `--screenshot on`, `--video on`) so the investigation is auditable.
+4. **Escalate to manual verification only as a last resort.**
+   - Manual testing is great for sanity checks but should be the exception, not the norm.
+   - Always record the exact steps and environment if you do perform a manual run.
+
+Keeping investigations scripted ensures regressions are caught by CI and reduces the effort required to share findings with the rest of the team.
+
+## Pull Request Checklist
+
+Before opening a pull request:
+
+- [ ] Ensure linting and tests pass locally.
+- [ ] Update documentation, examples, and configuration defaults when behaviour changes.
+- [ ] Include any new or updated Playwright tests/snapshots required to prove UI changes.
+- [ ] Provide clear reproduction steps and links to failing tests/logs in the PR description.
+
+We appreciate your contributions! If anything in this guide is unclear, open an issue so we can improve it.

README.md

@@ -276,6 +276,10 @@ The `examples/` folder contains end-to-end scenarios you can run as templates:
 
 Each example is annotated with the commands needed to reproduce the run, making it easy to adapt to your own workflows.
 
+## Contributing
+
+Interested in improving FlowDex? Read the [contribution guide](CONTRIBUTING.md) for setup instructions and the recommended debugging workflow (automated tests ➝ Playwright verification ➝ manual checks only when unavoidable). The guide also includes a pull-request checklist to keep changes ship-ready.
+
 ---
 
 © 2025 FlowDex. Generated 2025-10-23T00:31:05.007257Z