{
  "$type": "site.standard.document",
  "bskyPostRef": {
    "cid": "bafyreiffrug3lbkrtsqzxyiqco5r4wisbwi7zltwjuorxp5aqkq4365kva",
    "uri": "at://did:plc:pgryn3ephfd2xgft23qokfzt/app.bsky.feed.post/3mnzsyucttej2"
  },
  "path": "/t/it-s-the-architecture-stupid-why-prompt-engineering-won-t-fix-agents/175246?page=2#post_25",
  "publishedAt": "2026-06-11T16:12:18.000Z",
  "site": "https://discuss.huggingface.co",
  "textContent": "Yes, that is exactly the connection I had in mind.\n\nORCA seems to move cognitive operations out of the prompt and into reusable runtime capabilities. DESi is trying to add a gating layer before execution: is this method allowed for this task, given the available evidence, constraints, uncertainty, and required inputs?\n\nSo in a combined architecture, ORCA-like skills could be the executable capabilities, while DESi-style method validity gating decides when a skill, tool, retrieval step, clarification state, or model call is epistemically justified.\n\nIn other words: reusable skills are the “how”; method validity gating is the “whether” and “under which conditions.”",
  "title": "“It’s the Architecture, Stupid” — Why Prompt Engineering Won’t Fix Agents"
}