A multiple AI agents to research, create, and get summary for the video scripts from social mediaπ±
If you want to check the announcement for the list of the winner, go check it out in here Hack Another Day - Quira Agents
Script Forge is the multi agent video content planner to research, create, and polish video scripts from social media. The system primary employs on evaluation to ensure high quality and engaging content.
-
Multi-Agent Architecture
- ResearchAgent: Plans and conducts topic research
- ScriptWriterAgent: Creates initial video scripts
- PolishingAgent: Enhances scripts for engagement
- EvaluatorAgent: Assesses quality and provides feedback
-
Advanced Content Research
- Dynamic search query generation
- Web content extraction and processing
- ChromaDB-based vector storage for semantic search
- URL deduplication and content filtering
-
Intelligent Script Generation
- Platform-specific formatting (TikTok, YouTube Shorts, Instagram Reels)
- Automatic timestamp generation
- Engaging transitions and hooks
- Visual suggestions and hashtag generation
-
Quality Assurance
- Automated evaluation of research quality
- Script engagement scoring
- Improvement suggestions
- Multiple retry attempts for low-scoring outputs
- Get the API Key from Deepseek API DeepSeek API & SerpAPI SerpAPI
- Clone this repo
git clone https://github.com/daviddprtma/Script-Forge.git
cd Script-Forge
- Create & activate a virtual env:
python -m venv venv
source venv/bin/activate # On Windows: venv\Scripts\activate
- Install dependecies
pip install -r requirements.txt
- Create .env file with your API Keys:
DEEPSEEK_API_KEY=your_deepseek_api_key
SERPAPI_KEY=your_serpapi_key
-
Run the script:
python deepseek_agent.py
-
Follow the interactive prompts to specify:
- Video format (TikTok, YouTube Short, Instagram Reel)
- Video length
- Number of questions/items
- Topic
- Number of URLs per search query
- Number of search terms
-
The system will generate:
- Research data
- Initial script
- Polished script
- Evaluation reports
- Human-readable output
output/
βββ [Topic]_[Timestamp]/
β βββ debug/
β β βββ (Debug logs and raw API responses)
β βββ research/
β β βββ (Research data and evaluations)
β βββ script/
β β βββ (Initial and polished scripts)
β βββ evaluation/
β β βββ (Evaluation reports)
β βββ final_script.json
β βββ final_script.txt