velos

Explain and operate the `Memory.swift` autoresearch loop in `Autoresearch`. Use when asked what the autoresearch tooling is, how the optimization loop works, how to bootstrap it, how to kick off a run, or how to inspect its outputs and status.

Run and compare Memory.swift evaluation harness (`memory_eval`) profiles with correct cache behavior, repeatable commands, and metric summaries. Use when asked to rerun evals, benchmark retrieval/storage changes, compare profiles, diagnose inconsistent scores, or produce report paths from `Evals/runs`.

Use when adding, updating, or reviewing CodeMode.swift evaluation scenarios, deterministic eval tests, or Wavelike-backed LLM eval coverage in this repository.

Use when designing, generating, validating, or importing synthetic evaluation datasets for CodeMode.swift eval scenarios.