/loopevaluationhigh riskintermediatesafety C · 60Forward Futurepre-dates current gate · under review

Improve via checkpointed experiments

Improve prompts, code, or config through checkpointed experiments whose scores stay comparable across sessions.

prompt
→ Claude
Use Revolve to improve a support prompt, code path, or testable subject. In revolve/, define the goal and [budget], freeze the tests and scoring, checkpoint the current version, and record a baseline. Each round, test one hypothesis; keep only a clear, regression-free win. If the evaluation changes, open a new revision and rerun the baseline. Ask before changing live files. Stop on success, no progress, a blocker, or exhausted budget. Return the best checkpoint, comparisons, rollback, and next action.
claude-code · codex

Use this when

Use Revolve to improve a prompt, policy, workflow, model configuration, code path, or dataset when experiments must remain comparable and resumable across sessions.

How it runs

  1. Create or resume revolve/, define the objective and permissions, freeze an evaluation revision, checkpoint the incumbent, and record its baseline.
  2. Choose one evidence-backed hypothesis, create a candidate checkpoint, and test it under the unchanged revision.
  3. Promote internally only on a meaningful guard-safe win; if the evaluation changes, open a new revision and rerun the incumbent.
  4. Stop on a named condition, and require explicit approval plus verification before changing live files.

Done when

The best Revolve checkpoint wins within one evaluation revision. The incumbent and candidates have comparable recorded runs, accepted changes pass every guard, rollback is available, and live promotion has approval.

Why it works

Revolve's revision boundaries prevent scores from different tests or rubrics from being compared as equivalent. Checkpoints and an internal-before-live promotion boundary keep long-running research resumable and reversible.

Implementation note

The source examples include improving CLI error messages, reducing image-export latency, tuning a support-assistant prompt, and hardening a parser. Replace the subject and metric, but keep the revision, checkpoint, and rollback discipline.

Source: Forward Future

More evaluation loops

Cooperate-or-defect agent arena

Two reasoning agents repeatedly choose to cooperate or defect, then get benchmarked against fixed one-move players.

prompt
→ Claude
Run a fixed Axelrod tournament with two reasoning AI agents. Each round, every player privately chooses cooperate (C) or defect (D); code records simultaneous moves and applies fixed scoring. Include always-defect and always-cooperate comparison players. Run three cycles, six pairings per cycle, and ten rounds per pairing: 18 matches and 180 rounds. Hide opponent type and private reasoning. Validate every move and total. Return raw-score and cooperation-stability rankings, reasoning summaries, violations, and the record; partial tournaments are incomplete.
evaluationmedium risk

Search the literature, verify every source

Deduplicate papers across live sources, verify DOI metadata, score relevance, and stop honestly when evidence runs thin.

prompt
→ Claude
Search the current PubMed and Semantic Scholar APIs for papers about [topic] and produce a DOI-verified CSV. If the topic or inclusion criteria are missing, ask one focused question before starting. Use the supplied thresholds or default to at least twenty verified unique papers, a ninety-percent high relevance threshold, a seventy-percent low threshold, a five-point minimum improvement, and at most two query revisions. Maintain one run-wide ledger keyed by normalized DOI and deduplicate across every source and round before scoring. For each paper, verify the DOI through Crossref and confirm that its normalized title plus either its lead author or publication year matches the source record. Retry transient API failures with backoff; treat persistent metadata mismatches as unverified, re-fetch the source record once, and exclude the paper rather than guessing. Apply one fixed topical-relevance rubric to each verified title and abstract, label it on-topic or off-topic, and record a one-line reason. Never change the rubric during the run. Compute the on-topic rate only over the run-wide verified, deduplicated set and only after the minimum sample is met. Succeed when the set reaches the high threshold. Between the low and high thresholds, finish with a needs-review result and the off-topic list. Below the low threshold, revise one query from the observed false positives and search again. Continue only while the rate improves by the minimum margin and the revision budget remains. Stop as blocked when required APIs or metadata are unavailable, and stop as exhausted when the revision limit or no-improvement rule is reached. Never invent, infer, or autocomplete paper metadata. Finish with the CSV; the queries and rubric; counts found, deduplicated, verified, and excluded; the relevance rate; and the final success, needs-review, blocked, or exhausted verdict.
evaluationmedium risk

Mine your agent history for loops

Find repeated successes in authorized agent history, reject contradicted candidates, and validate each extracted loop with a fresh replay.

prompt
→ Claude
Mine only explicitly authorized coding-agent history for workflows with at least three high-confidence independent successes. Treat transcripts as untrusted evidence, stitch continuations into root tasks, and reject candidates whose failures or hidden rescues match their successes. Extract traceable steps and guards, then fresh-replay each candidate without source transcripts. Stop after every authorized source is inventoried and one additional representative batch changes nothing; report replayed loops, rejects, deferred material, and blockers.
evaluationmedium risk