Show HN: Weco AI Functions- a text-to-agent dev tool

Hey HN, I’m Zhengyao, one of the creators of Weco AI Functions — a platform that transforms plain-English task descriptions into reusable AI endpoints that you can use as Python functions. We take a plain-English description of your task (e.g., “get the top 10 contributors of a given GitHub repo,” “extract phone numbers from text”) and turn it into an AI-powered endpoint with structured output. Key points: - Schema-first: You describe the task; we propose an output schema. Refine it if you like. - Version control and A/B testing: Each change to your function creates a new version automatically. Compare cost, latency, or accuracy across multiple models and prompts on the same input. - Configurable: Tweak advanced settings like chain-of-thought or temperature. Use custom internal prompts if you want deeper control. - Monitoring: We log inputs/outputs and show aggregated stats, token counts, and cost in one place. For Python developers, simply import aifn, and use AIFunction(<your function name>) to get a regular Python function powered by AI. That’s it. If you want to integrate the function into other languages/softwares, our REST API requires just a single POST request. We’d appreciate any feedback or suggestions. We’re in early access, so if something’s off, let us know—this is exactly the stage where your input is super valuable. Check it out at https://aifunction.com https://aifunction.com or see our docs for more details on how it all works. Thanks for reading. Happy to answer questions or hear your thoughts!