← Operator PowersCONTENT HISTORYWHAT CHANGED · RULE-BASED ANALYSIS
Update to Operator Powers
Snapshot Sep 30, 2026 · 22:54 UTC · version 1.0.1
Collection source: not recorded for this historical snapshot.
First saved snapshot
No earlier snapshot is available to establish a change.
Compare saved observations
Download comparison JSONFull technical diff · 0 changed fields
Full snapshot data
{
"name": "big-model-smell",
"description": "Simplify an oversized prompt, plan, workflow, document, system, or build when its complexity, context, automation, or model cost is not earning its keep.",
"included_files": [
{
"relative_path": "agents/openai.yaml",
"size_in_bytes": 226
}
],
"skill_md_contents": "---\nname: big-model-smell\ndescription: Simplify an oversized prompt, plan, workflow, document, system, or build when its complexity, context, automation, or model cost is not earning its keep.\n---\n\n# Big Model Smell\n\nFind where complexity, context, automation, or model capability exceeds the actual job, then simplify without losing the required outcome.\n\n## Job contract\n\n- Owns: simplification of one supplied artifact, workflow, prompt, system, or plan.\n- Does not own: retrospective analysis across recent sessions (`operator-audit`).\n- Finished deliverable: complexity diagnosis, smallest viable design, migration steps, and verification criteria.\n\n## Smells to inspect\n\n- One giant prompt trying to perform unrelated jobs.\n- A frontier model used for deterministic extraction or formatting.\n- Too many agents, tools, files, approval steps, or moving parts.\n- Context loaded “just in case” rather than for a decision.\n- An automation that costs more to maintain than the task costs to do.\n- A document or system no operator can confidently change.\n\n## Workflow\n\n1. State the outcome and non-negotiable constraints.\n2. Map every component to the outcome it supports.\n3. Mark components as essential, useful, unproven, duplicate, or decorative.\n4. Identify the cheapest adequate level for each step: rule, template, small model, large model, human judgment, or automation.\n5. Design the smallest version that preserves safety and quality.\n6. Give a reversible migration order and a test that proves nothing important was lost.\n\n## Output\n\n```markdown\n# Complexity Review\n\n## Required Outcome\n[What must remain true]\n\n## Big-Model Smells\n[Evidence and cost of each]\n\n## Keep, Simplify, Remove, Test\n[Decision table]\n\n## Smaller Working Design\n[New flow]\n\n## Safe Migration\n[Reversible sequence]\n\n## Verification\n[Tests and stop conditions]\n```\n\n## Guardrails\n\n- Do not simplify away approval, privacy, security, accessibility, or recovery controls.\n- Do not assume a cheaper model is adequate; define a representative test.\n- Never modify or delete the original artifact without explicit approval.\n"
}SHA-256: a4b39020fc999475369813fbdee02050fdcc2828530088b2bd41219829c939da