Rifft
Site offlineAI agent debugger for CrewAI, AutoGen and LangGraph
Details
- Categories
- AIDeveloper ToolsAnalytics & Monitoring
- Use Cases
- AI AgentsTesting & QABug Tracking
- Target Audience
- AI EngineersAI DevelopersSoftware Developers
- Pricing
- Freemium
- Platforms
- Web
About Rifft
You can debug multi-agent pipeline failures faster with Rifft. This tool helps you trace root causes and classify failures using the MAST taxonomy. You can also fork and replay any specific handoff without needing to restart your entire pipeline, saving you significant development time.
Product Insights
Rifft is a web-based debugging platform designed specifically for multi-agent AI frameworks such as CrewAI, AutoGen, and LangGraph. It leverages the MAST taxonomy to help developers trace failures and re-execute specific pipeline handoffs.
- Supports specialized multi-agent frameworks including CrewAI, AutoGen, and LangGraph.
- Provides root cause classification using the standardized MAST taxonomy.
- Allows forking and replaying specific agent handoffs to avoid full pipeline restarts.
- Offered as a freemium web application for streamlined accessibility.
Ideal for: Rifft is ideal for AI Engineers and Software Developers who need to isolate and resolve failures in complex multi-agent AI pipelines.
Product Video
Watch a video demo of Rifft.
Screenshots
Reviews (0)
No reviews yet. Be the first to rate this product!
Comments (3)
Debugging multi-agent frameworks like CrewAI and LangGraph is genuinely painful right now. A dedicated debugger for these systems fills a real gap that everyone building with them feels.
Interesting idea! I like it
Built after too many hours debugging agent failures with no idea why. Rifft traces every handoff, classifies failures automatically, and lets you replay broken runs.