Today we are publishing our first collaborative research with @Eon_io_
We gave an Eragon agent a readable guide to a company's systems, then let GEPA rewrite that guide from the agent’s own failed traces.
GPT-5.6 Luna is the cheapest model we tested, and cold it was the weakest of the GPT and Claude models: 7/35.
Post GEPA: 29/35.
Show more
Eragon now has its own drink, available now exclusively at @UseCorgi Cafe in SF. Introducing the Eragon Noir, a sesame paste infused latte.
Dark, complex, but surprisingly smooth -- kind of like deploying Eragon agents in production.
Swing by Corgi Cafe and try it. Open 24/7. Just like your Eragon agent.
Show more
