Explore Prompts

Design a loop to improve job scheduling: fairness, retries, visibility timeouts, and backpressure. Require load testing and chaos testing in staging.

Create an agent that validates docs examples by executing them in a sandbox, checks links, and ensures version correctness. Propose fixes with evidence.

Design a plan to self-improve ranking/retrieval safely: offline eval sets, interleaving/AB tests, bias checks, and rollback on metric drops.

Create an agent to tune caches: TTLs, invalidation, and hit rates. Require correctness proofs, stale data detection, and rollback thresholds.

Design an agent that improves ETL reliability: retries, idempotency, checkpoints, and backfills. Require correctness validation and cost constraints.

Create a strict policy for edits to auth code: mandatory human review, extra tests, formal checks, and staged rollout. Include “never auto-merge auth changes.”

Design an agent that tunes security scanners (SAST rules, allowlists) based on confirmed findings and false positives. Require approvals for any rule weakening.

Create a blueprint to improve build scripts safely: hermeticity, caching, parallelism, and reproducibility. Require staged validation and rollback.

Design a multi-repo agent that only edits repos it’s authorized for, respects codeowners, and uses per-repo policies. Include cross-repo dependency coordination rules.

Create a loop where production failures and near-misses become new eval tests. The agent should propose test additions with minimal reproductions and acceptance criteria.

Design a process to adjust safety filters based on measured false positive/negative rates. Require evaluation sets, human review, and rollback if harm risk rises.

Design governance for self-improving systems: approval rules, quarterly audits, access reviews, and incident drills. Include “who can change the agent” controls.