PromptForge - One CLI. Ten agents. One perfect prompt.

A production-ready Python CLI that runs a multi-agent sequential pipeline via Ollama to generate self-contained master prompts optimized for AI code editors (Cursor, Claude Code, Windsurf, GitHub Copilot). Turn an app idea into a copy-pasteable prompt containing architecture diagrams, implementation plan, business impact analysis, bottleneck detection, security audit, AI strategy recommendations, and a technology radar — all in one shot.

Hey PH community 👋 I'm Bharath R, builder of PromptForge. Why I built this: I kept losing hours stitching together architecture diagrams, security checklists, and implementation plans just to get a good prompt for Cursor/Claude. So I automated the pipeline. What makes it different: • 10 specialist agents (Architect, Security Auditor, Business Analyst, etc.) that chain context automatically • 100% local via Ollama—no API keys, no data leaves your machine • Outputs are editor-ready: copy-paste into Cursor, Windsurf, or Copilot and start coding I'd love your feedback on: Which agent output is most valuable to you? (diagrams, ROI, bottleneck scan, etc.) Would you use this solo, or integrate it into a team workflow? What's missing for your stack? 🔗 GitHub: https://github.com/BharathR1703/... 🎬 demo:

Built to ship faster, not prompt harder. 🚀

PromptForge - One CLI. Ten agents. One perfect prompt.

Replies