Explore Prompts

Create a threat-modeling step: identify assets, attackers, and abuse cases for each PR. Include automated checks and a human escalation threshold.

Design a process where the agent checks a proposed change against official documentation versions, release notes, and migration guides. Require citations for claims.

Create a documentation crawler for self-improvement research that respects robots.txt, rate limits, and TOS. Include caching, dedupe, and citation storage.

Design an improvement loop keyed to SLOs: latency, error rates, and quality metrics. Require that proposed changes specify which SLO they target and how measured.

Create a backlog system: bugs, tech debt, performance, and security items. Include scoring by user impact, risk, and effort, and automatic issue creation via MCP.

Design a patch strategy that prefers minimal diffs: isolate behavior changes, avoid formatting churn, and include clear intent comments. Add a “diff size budget.”

Define guardrails that block large architectural changes unless an explicit human-approved RFC exists. Include heuristics to detect “too big” diffs.

Create an agent that builds a “system map”: modules, ownership, dependencies, and critical paths. Use it to constrain edits to low-risk areas first.

Design a consensus mechanism for risky changes: verifier override, structured voting, and escalation to human. Include disagreement logs and rationale storage.

Build an eval harness for self-edits: deterministic tool mocks, seeded randomness, replayable runs, and stored artifacts for auditing decisions.

Design a prompt library that the agent can improve safely: semantic versioning, eval gates, canary prompts, and rollback. Include prompt linting rules.

Create a sandbox spec: network isolation, filesystem constraints, resource limits, and safe temp dirs. Include how MCP tool wrappers enforce the sandbox.