Delete README_SPACE.md

README_SPACE.md

@@ -1,164 +0,0 @@
-# Multi-Agent Content Generation System - Hugging Face Space
-
-This is a Hugging Face Space deployment of the multi-agent content generation system. The system orchestrates seven specialized AI agents to collaboratively produce high-quality written content.
-
-## Features
-
-- **7 Specialized Agents** - Showrunner, Story Editor, Cultural Consultant, Lead Writer, Dialogue Specialist, Comedy Writer, and Proofreader
-- **API-First Architecture** - RESTful endpoints for all operations
-- **OpenRouter Integration** - Powered by `poolside/laguna-m.1:free` model
-- **Hugging Face Integration** - Automatic upload of outputs to dataset
-- **Persistent Storage** - All agent outputs saved as JSON
-
-## Setup
-
-### 1. Create the Space
-
-Go to [Hugging Face Spaces](https://huggingface.co/spaces) and create a new Space:
-- **Name:** `multi-agent-system` (or your preferred name)
-- **License:** MIT (or your choice)
-- **Space SDK:** Docker
-- **Visibility:** Private (recommended for security)
-
-### 2. Add Secrets
-
-In your Space settings, add the following secrets:
-
-#### `OPENROUTER_API_KEY`
-Your OpenRouter API key for accessing the `poolside/laguna-m.1:free` model.
-- Get it from: https://openrouter.ai/keys
-
-#### `HUGGINGFACE_TOKEN`
-Your Hugging Face user access token with write permissions.
-- Get it from: https://huggingface.co/settings/tokens
-- Must have write access to the `factorstudios/Pipeline` dataset
-
-### 3. Configure Environment Variables
-
-Optional environment variables (set in Space settings if needed):
-
-- `HUGGINGFACE_DATASET` - Target dataset (default: `factorstudios/Pipeline`)
-- `OPENROUTER_BASE_URL` - OpenRouter API endpoint (default: `https://openrouter.ai/api/v1`)
-- `MODEL_NAME` - Model to use (default: `poolside/laguna-m.1:free`)
-- `LOG_LEVEL` - Logging level (default: `INFO`)
-
-## Usage
-
-### Health Check
-
-```bash
-curl https://your-space-url/health
-```
-
-### Execute Full Pipeline
-
-```bash
-curl -X POST https://your-space-url/api/v1/pipeline/execute \
-  -H "Content-Type: application/json" \
-  -d '{
-    "user_brief": "A comedic episode about...",
-    "season_arc_document": "Season context...",
-    "character_bible": "Character definitions...",
-    "world_building_document": "World context...",
-    "character_voice_guide": "Voice definitions...",
-    "style_guide": "Style reference...",
-    "continuity_log": "Continuity tracking...",
-    "hook_brief": "Optional hook..."
-  }'
-```
-
-### API Documentation
-
-Visit `https://your-space-url/docs` for interactive API documentation.
-
-## API Endpoints
-
-### Pipeline Management
-- `GET /health` - Health check
-- `POST /api/v1/pipeline/execute` - Execute full pipeline
-- `GET /api/v1/pipeline/status/{run_id}` - Get pipeline status
-
-### Individual Agents
-- `POST /api/v1/showrunner/generate_directive` - Generate directive
-- `POST /api/v1/story_editor/generate_outline` - Generate outline
-- `POST /api/v1/cultural_consultant/review_outline` - Review outline
-- `POST /api/v1/lead_writer/write_script` - Write script
-- `POST /api/v1/dialogue_specialist/polish_dialogue` - Polish dialogue
-- `POST /api/v1/comedy_writer/add_humor` - Add humor
-- `POST /api/v1/proofreader/final_qc` - Final QC
-
-## Architecture
-
-```
-FastAPI Server (Port 7860)
-    ↓
-Pipeline Orchestrator
-    ↓
-7 Agent Modules
-    ↓
-OpenRouter API (poolside/laguna-m.1:free)
-    ↓
-Hugging Face Dataset Upload
-```
-
-## Security
-
-- **Secrets Management** - API keys stored securely in HF Secrets
-- **No Hardcoded Credentials** - Environment variables loaded at runtime
-- **HTTPS Only** - All communication encrypted
-- **Private Space** - Recommended for production use
-
-## Performance
-
-Typical pipeline execution time: **10-15 minutes**
-
-Each agent processes sequentially through the OpenRouter API:
-1. Showrunner: ~40 seconds
-2. Story Editor: ~150 seconds
-3. Cultural Consultant: ~130 seconds
-4. Lead Writer: ~140 seconds
-5. Dialogue Specialist: ~50 seconds
-6. Comedy Writer: ~120 seconds
-7. Proofreader: ~40 seconds
-
-## Storage
-
-- **Local Storage** - Agent outputs saved in Space filesystem
-- **Hugging Face Dataset** - Final outputs uploaded to `factorstudios/Pipeline`
-- **Metadata** - Pipeline execution metadata also uploaded
-
-## Troubleshooting
-
-### 401 Unauthorized
-- Check `OPENROUTER_API_KEY` is set correctly in Secrets
-- Verify key has access to `poolside/laguna-m.1:free` model
-
-### 403 Forbidden (Hugging Face)
-- Check `HUGGINGFACE_TOKEN` is set correctly in Secrets
-- Verify token has write access to dataset
-- Confirm dataset exists: `factorstudios/Pipeline`
-
-### Timeout Errors
-- Pipeline execution can take 10-15 minutes
-- Increase request timeout to 600+ seconds
-- Check Space logs for details
-
-### Out of Memory
-- Spaces have limited memory
-- Consider splitting pipeline into separate requests
-- Use individual agent endpoints instead of full pipeline
-
-## Logs
-
-View Space logs in the Hugging Face Space interface for debugging.
-
-## Support
-
-For issues or questions:
-1. Check the logs in your Space
-2. Review API documentation at `/docs`
-3. Verify all Secrets are configured correctly
-
-## License
-
-This project is part of Factor Studios' multi-agent content generation initiative.