@paircoder

we spent 12 months teaching Python to babysit Claude Code 🐍 enforcement gates · 200+ commands · 7,500+ tests · 98% core coverage stop hoping. start enforcing.

 paircoder.ai
paircoder.ai
Joined August 2020
  • Tweets702
  • Following44
  • Followers81
1Media
paircoder retweeted
Juan Suarez@jlsuarezFeb 21
“Claws are now a new layer on top of LLM agents.” That’s how Andrej Karpathy framed what many teams are starting to see in practice. Agents alone are no longer enough. We moved from prompts to chatbots to agents. Now a control layer is forming above them.
2
1
1
92
paircoder retweeted
aioan@AIoan_Feb 20
Replying to @paircoder
btw i dropped you a fllw cuz I think you are building cool stuff
2
3
53
My recommendation rn is that you have AI review its own code at least 3 times before a human review. Once at the end of the current turn. Once before you submit the PR. Once after PR submission. If you do that much you will weed out a lot of dumb mistakes from the agent.
28
10
4
122
8,883
Load more