{
  "$type": "site.standard.document",
  "bskyPostRef": {
    "cid": "bafyreidf5cx33fcuf72y6mzda5bm7b4nvgjrybq3cvrqraicbdl7fcdisi",
    "uri": "at://did:plc:lk3jfj3zq4k4wxnk474axylu/app.bsky.feed.post/3ml6rcxhyiou2"
  },
  "path": "/t/gpt-massive-anxiety-stress/1380366#post_4",
  "publishedAt": "2026-05-06T12:47:25.000Z",
  "site": "https://community.openai.com",
  "textContent": "I disagree. Been coding w/ codex since inception. This behavior is exactly as OP described! I am beside myself with all the arbitrary editing, it now deletes unrelated code blocks, out of scope with instructions!\n\nin spite of prompts, and ignores REPEATED REQUESTS 10+ TIMES!?\n\nLOOKS LIKE WE’RE IN A TAILSPIN TO MEDIOCRITY…\n\nCONFIRMED\n\n  * The model kept steering away from the user’s actual objective.\n  * It substituted soft artifacts, outreach, summaries, and manual capture for executable proof.\n  * It repeatedly failed to preserve the core requirement: automated, runner-recorded, sealed lower-boundary evidence.\n\n",
  "title": "Gpt = massive anxiety+stress"
}