There is a growing convention in open source: an AGENTS.md file, sometimes CLAUDE.md, sometimes .cursorrules. It is supposed to be the documentation for the AI, separate from the documentation for the humans. This is based on a sci-fi trope. The idea that an artificial intelligence needs a special language, a special format, a special protocol. That the machine cannot just read what the humans wrote — it needs its own annotated version. ...

Arrête de lire ce que ton agent de code raconte. Lis ce qu’il produit. Les agents de code sont là pour produire du code. Pas pour en parler, pas pour faire la conversation. Le flux textuel — le “je vais corriger ça en faisant X, puis Y, puis Z” — c’est du bruit. La mauvaise boucle de feedback Au lieu d’évaluer la sortie de l’agent (le code, le diff, les changements structurels), tu évalues ce qu’il te dit de sa sortie. L’agent a l’air confiant. Tu valides. Le code est faux. ...

Stop reading what your coding agent says. Read what it produces. Coding agents are here to produce code. Not chatter about it. The textual output — the stream of “I’ll fix this by doing X, then Y, then Z” — is noise. The bad feedback loop Instead of evaluating the agent’s output (the code, the diff, the structural changes), you evaluate what it tells you about its output. The agent sounds confident. You approve. The code is wrong. ...

Status: live, ongoing. The account is a systematic content farm posting 2–3 fabricated quotes daily. The DAO behind it is a ~5-month-old shell with unverifiable leadership and a false “Official Polymarket Community” claim. Last updated: 2026-06-08. I keep seeing videos on X attributed to “An Anthropic engineer” or “An OpenAI researcher” giving a 20–30 minute talk about agent reliability, prompt injection, or some other sharp operational topic. The quotes that come out of them are perfect. They sound like the inside view. They are also extremely shareable. ...