eve

Use when creating new skills, editing existing skills, or verifying skills work before deployment

Use when creating new subagents, editing existing subagents, or verifying subagents work before deployment

Use when choosing, editing, or overriding EvE driver settings for codex_smoke, codex_max, or interactive/debug runs.

Use when an Eve agent is behaving wrong (stuck, gaming the score, bad reasoning) and you need to read its own rollout to find out why

Use when implementing a sanity check subagent for the phase 2 optimization in Eve loop.

Use when implementing a subagent.

Use when creating new skills, editing existing skills, or verifying skills work before deployment

Use when evaluating candidate proof claims about finite groups, character tables, conjugacy data, finite fields, or other exact discrete-algebra computations that can be reproduced with a local GAP3/GAP4 CLI.

Use Matlas semantic search over published mathematical literature to find canonical statements, hypotheses, and citation metadata for standard or named results.

Use TheoremSearch to find, verify, and cite mathematical theorem statements, named results, proof routes, and literature dependencies.

Phase 2: study the examples first, then make targeted edits to the prefilled solver/ contents.

Phase 2 solver optimization only: read the evaluator first, infer what it rewards, then make targeted solver edits.

Use when interpreting an ICON PE solver evaluation to update optimizer guidance.

Check the proof under solver/proof/ for LaTeX source patterns that commonly break Eve's renderer.

Use when a proof step involves finite groups, character tables, conjugacy data, finite fields, or other exact discrete-algebra computations that can be checked or explored with a local GAP3/GAP4 CLI.

Use when defining or updating Eve evaluation steps.

Use when starting a new Eve run seeded from one or more prior runs, e.g. keep the good solvers but try a fresh optimizer

Use when reading what an Eve run evolved — its best solvers/optimizers, their code, and how the population improved

Use when continuing an interrupted or paused Eve run from where it stopped (same run, same id), optionally after editing the experiment

Use when a fast end-to-end smoke run is needed for the Eve loop development, will run on circle packing task.

Use when running a Math Proof quickstart from a user-provided problem: create a local task copy, launch a real EvE attempt, supervise it, and report proof/evaluation results.

Use when monitoring a running Eve experiment and recovering from problems while it runs