@garrytan: My friend @finbarr says: It's like code as memory. You work with your agent in a non deterministic way to figure out ho…
Summary
Garry Tan discusses the concept of "code as memory" for AI agents, suggesting they generate executable scripts for new tasks and reuse them for efficiency.
Similar Articles
@garrytan: My simple secret to agentic coding https://forbes.com/sites/josipamajic/2026/04/12/the-yc-chief-who-codes-10000-lines-a…
Garry Tan of Y Combinator shares his 'thin harness, fat skills' agentic coding framework, while the accidental leak of Claude Code's source code reveals the complex scaffolding behind AI coding agents.
@garrytan: https://x.com/garrytan/status/2061454423034110372
Garry Tan argues that developers are over-engineering with excessive code when using AI agents; instead, they should trust the model and build minimal, instruction-based software, exemplified by his open-source project GStack.
@AlperTheKing: Garry Tan's GBrain makes the memory write path the reliability boundary for agent systems, because useful context must …
Garry Tan's GBrain introduces an agent memory system that treats the memory write path as the reliability boundary, using Markdown as source of truth with Postgres and pgvector for retrieval, and exposing operations via CLI and 30+ MCP tools.
@garrytan: https://x.com/garrytan/status/2054064931515855118
Garry Tan argues that AI coding agents like Claude Code and Codex have changed software engineering by making high test coverage affordable, creating a 'complexity ratchet' that ensures code quality improves over time without sacrificing speed.
@davidNbreslauer: Gbrain by @garrytan is so good, I've now moved by brain onto a dedicated server, that Codex, Claude, OpenClaw, and Herm…
Garry Tan introduces GBrain, an open-source AI agent memory layer that provides synthesis, graph traversal, and gap analysis for coding agents, demonstrating significant improvement over traditional retrieval methods.