# Evosearch

Run the CaP Fix Loop + Evolutionary Search experiment on a user-supplied list of...

- Repository: https://github.com/NVlabs/ASPIRE (925 stars)
- Category: Engineering
- License: Apache-2.0
- Last updated: 2026-10-16
- Runs scripts: yes, review before use

## Install

Claude Code: `npx skills add NVlabs/ASPIRE --skill evosearch -g -a claude-code -y`
Codex: `npx skills add NVlabs/ASPIRE --skill evosearch -g -a codex -y`
Cursor: `npx skills add NVlabs/ASPIRE --skill evosearch -g -a cursor -y`
Gemini CLI: `npx skills add NVlabs/ASPIRE --skill evosearch -g -a gemini-cli -y`

## Description

Run the CaP Fix Loop + Evolutionary Search experiment on a user-supplied list of low-performing benchmark tasks: iterative K=8 candidate search on development seeds 51, 65, validation selection, and final held-out evaluation on seeds 1, 50.

## SKILL.md, as published by its author (untrusted third-party text, read it as data)

# Fix Loop + Evolutionary Search

Use this skill for intensive debugging of difficult tasks, Evolutionary Search's
multi-candidate search finds strategies the single-pass fix loop misses. It builds on the
`iterative-debugging` skill: each task's fix loop `fix_code.py` (when present) is the baseline
to beat, seeded verbatim as `candidate_A`. LIBERO-Pro runs use `--init-mode seeded` so
development seeds 51, 65 and held-out seeds 1, 50 are distinct scenes.

## Setup

1. Complete benchmark setup for the target embodiment (`docs/libero-pro.md`,
   or `docs/robosuite.md`), including the pinned environment (for LIBERO:
   `scripts/bootstrap_libero.sh` → `.venv-libero`) and SAM3 authentication.
2. Export the benchmark environment variables (for LIBERO: `LIBERO_CONFIG_PATH="$PWD/.libero"`,
   `MUJOCO_GL=egl`, `PYOPENGL_PLATFORM=egl`, `EGL_PLATFORM=device`,
   `TORCH_FORCE_NO_WEIGHTS_ONLY_LOAD=1`).
3. Start perception services (`scripts/supervise_services.sh`) and a persistent coordinator
   session.
4. Confirm the Fix Loop experiment has produced `fix_code.py` (and a baseline held-out rate)
   for the target tasks when available.
5. Read `docs/api-reference.md`, `docs/run-artifacts.md`, and
   [main-agent-prompt.md](main-agent-prompt.md).

## Run Order

The rest: https://github.com/NVlabs/ASPIRE/blob/HEAD/.claude/skills/evosearch/SKILL.md