A curated selection of real conversations with AI coding agents like
Claude Code, OpenCode, and Codex CLI. Not every session makes this list.
These are the ones where i've felt like it's genuinely done something
impressive or extremely long running tasks. LLM's are truly powerful with
the proper harness, environment and tools around it.
Each entry is a condensed, human-readable retelling of the session.
Sensitive details (credentials, internal URLs, proprietary logic) are
redacted. The focus is on the technical approach and what made it work.