Reading up on Clawd
5 deep · digging since jan 21
- In defense of not reading the code
Ben Shoemaker argues that for most product code, reading line-by-line is an inefficient verification method, advocating instead for layered automation, specs, tests, and harness design around AI-generated code.
- The creator of Clawd: "I ship code I don't read" - YouTube
Peter Steinberger, creator of Clawd, made over 6,600 commits in January and says he ships code without reading it.
Takes
CRITICAL: EVERYONE USING CLAWDBOT SHOULD RUN THIS PROMPT RIGHT NOW:By default, the 2 best Clawd memory features are turned OFFRunning this 1 prompt will immediately stop your AI from getting confused in between compactions/sessions:Prompt:"Enable memory flush before…
@AlexFinn
This demo is the craziest thing you’ll see today. Full stop.Watch Clawd SIGN UP for a Reddit account completely autonomously with its own email account (thru @agentmail) + web browser.The next six months are going to be wild. pic.twitter.com/B2Jh5BehJj
@mattshumer_
Yeah this was 1,000% worth it. Separate Claude subscription + Clawd, managing Claude Code / Codex sessions I can kick off anywhere, autonomously running tests on my app and capturing errors through a sentry webhook then resolving them and opening PRs... The future is here. https://t.co/dKxOYZfYAl
@nateliason