Hey Product Hunt! 👋 David here, CEO of Expanso.

We built skills.expanso.io because we kept seeing the same problem: every team rewriting the same data processing logic from scratch — PII removal, log parsing, schema enforcement, format transforms. Each time, from zero. Each time, subtly broken.

So we made them reusable. Each skill is a self-contained data pipeline recipe that runs on Expanso Edge, and is directly callable from your agentic framework or directly, no AI needed! Think of it like npm packages, but for data processing:

🔒 remove-pii — Strip sensitive fields before your AI agent ever sees them
📊 parse-logs — Turn 1,000 raw log lines into 1 structured digest (99.9% reduction)
🌍 cross-border-gdpr — Automatically route data based on jurisdiction
🔐 encrypt-data — End-to-end encryption enforced at the pipeline level
📡 fan-out-kafka / fan-out-s3 — Distribute processed data to multiple destinations simultaneously

172 skills available today. All open source. All composable.

The core idea: your AI agents should never touch raw data. Skills filter at the source, enforce least privilege, and give agents only what they need — nothing more.

Self-hostable, works at the edge, plays nicely with LangChain, CrewAI, AutoGPT, and any agent framework.

Try it right now: https://exso.cloud/try

We'd love your feedback — what data processing patterns are you rebuilding repeatedly? Drop a comment and we'll prioritize adding them.

Thanks for the support 🙏

Expanso Skills

Data processing recipes for AI agents

Data processing recipes for AI agents