ar-run

原文涵盖的主题

• /ar-run — Single Experiment Iteration
• Usage
• What It Does
• Step 1: Resolve experiment
• Step 2: Load context
• Step 3: Decide what to try
• Step 4: Make ONE change
• Step 5: Commit and evaluate

Run exactly ONE experiment iteration: review history, decide a change, edit, commit, evaluate. If no experiment specified, run python {skillpath}/scripts/setupexperiment.py --list and ask the user to pick. Review results.tsv: What changes were kept? What pattern do they share? What was discarded? Avoid repeating those approaches. What crashed? Understand why. How many runs so far? (Escalate strategy accordingly) Strategy escalation: Runs 1-5: Low-hanging fruit (obvious improvements) Runs 6-15: Systematic exploration (vary one parameter) Runs 16-30: Structural changes (algorithm swaps) Runs 30+: Radical experiments (completely different approaches) Edit only the target file specified in config.cfg. Change one thing. Keep it simple. Read the script output. Tell the user: KEEP: "Improvement! {metric}: {value} ({delta} from previous best)" DISCARD: "No improvement. {metric}: {value} vs best {best}. Reverted." CRASH: "Evaluation failed: {reason}. Reverted." After every 10th experiment (check results.tsv line count), update the Strategy section of program.md with patterns learned. ONE change per iteration. Don't change 5 things at once. NEVER modify the evaluator (evaluate.py). It's grou

这里只展示节选，请以完整原文和当前要求为准。