Spaces:
Sleeping
Sleeping
| title: FlowPilot | |
| emoji: "📬" | |
| colorFrom: blue | |
| colorTo: green | |
| sdk: docker | |
| app_port: 7860 | |
| short_description: Gmail-first small business workflow automation with FastAPI. | |
| # FlowPilot | |
| FlowPilot is a Gmail-first automation layer for small business owners. The project includes: | |
| - A FastAPI backend with analysis, workflow suggestion, workflow build, deploy, upload, status, and escalation endpoints. | |
| - A primitive-based workflow engine that compiles and executes JSON workflows. | |
| - A Chrome extension scaffold that injects a sidebar into Gmail and walks through the onboarding flow. | |
| - Lightweight demo storage and tests for the core owner flow. | |
| ## Project Structure | |
| ```text | |
| backend/ | |
| extension/ | |
| tests/ | |
| Dockerfile | |
| README.md | |
| ``` | |
| ## Local Development | |
| Install dependencies and run the backend: | |
| ```bash | |
| pip install -r backend/requirements.txt | |
| uvicorn backend.main:app --reload | |
| ``` | |
| Create a `.env` in the repo root first: | |
| ```bash | |
| nano .env | |
| ``` | |
| Set these values when you want live Groq responses: | |
| ```bash | |
| AI_PROVIDER=groq | |
| GROQ_API_KEY=your_groq_key | |
| GROQ_MODEL=llama-3.3-70b-versatile | |
| GROQ_BASE_URL=https://api.groq.com/openai/v1 | |
| GROQ_TIMEOUT_SECONDS=8 | |
| ``` | |
| If `GROQ_API_KEY` is blank, FlowPilot falls back to local deterministic logic so the app and tests still work. | |
| ## Hugging Face Space | |
| This repo is ready for a Docker Space. The container listens on port `7860`, and the metadata at the top of this README tells Hugging Face to treat it as a Docker app. | |
| Typical flow: | |
| ```bash | |
| git clone https://huggingface.co/spaces/technophyle/flow-pilot | |
| cd flow-pilot | |
| ``` | |
| Copy this project into the cloned Space repo, then push: | |
| ```bash | |
| git add . | |
| git commit -m "Switch FlowPilot to Hugging Face Space with Groq" | |
| git push | |
| ``` | |
| In the Hugging Face Space settings, add these secrets: | |
| - `GROQ_API_KEY` | |
| Optional Space variables: | |
| - `AI_PROVIDER=groq` | |
| - `GROQ_MODEL=llama-3.3-70b-versatile` | |
| - `GROQ_BASE_URL=https://api.groq.com/openai/v1` | |
| - `GROQ_TIMEOUT_SECONDS=8` | |
| - `ANALYZE_WITH_AI=false` | |
| For GitHub Actions auto-deploys, add these repository settings: | |
| - Secret: `HF_TOKEN` | |
| - Variable: `HF_SPACE_REPO=technophyle/flow-pilot` | |
| ## Extension | |
| Load `extension/` as an unpacked Chrome extension. The sidebar currently points to `http://localhost:8000/api` for local work, so update that base URL when you connect it to your hosted Space backend. | |
| ## Notes | |
| - Backend storage is currently lightweight demo storage suitable for iteration. | |
| - Groq is used for live AI calls through its OpenAI-compatible chat completions API. | |
| - The `/api/analyze` endpoint can still stay fast and deterministic when `ANALYZE_WITH_AI=false`. | |