Til 2025-07-03

• DeepEval is a framework for testing LLM results.

• Everything in AI architecture is overwhelming, but all they are easier to understand if we think as how to work as a team.

• Hasura team is awesome, reading through the new promptQL blew my mind.

---

Today read

• https://blog.langchain.com/context-engineering-for-agents/?ref=dailydev

• https://www.anthropic.com/engineering/building-effective-agents?ref=blog.langchain.com

• https://www.dbreunig.com/2025/06/26/how-to-fix-your-context.html

• https://cognition.ai/blog/dont-build-multi-agents#principles-of-context-engineering

• https://promptql.io/docs/capabilities/