📡 MCP Radar

gatordevin/avo

Open reproduction of NVIDIA's AVO paper (arXiv:2603.24517): evolutionary search where an autonomous coding agent IS the variation operator — Vary(P)=Agent(P,K,f). Runs on the Claude Code or Codex sess

3 stars
1 forks
Python
momentum ▲ 7.0
created 2026-08-22
on radar since 2026-08-26
2048claude-codecodexcoding-agentsevolutionary-algorithmsexpectimaxkernel-optimizationllm-agentsmcpmcp-serverpaper-implementationreproducibility
View on GitHub ↗

About avo

An open reproduction of AVO: Agentic Variation Operators for Autonomous Evolutionary Search (Chen, Ye, Xu et al., NVIDIA, 2026), runnable on a laptop.

Classical evolutionary search, and the LLM-augmented systems that followed it, decompose the variation operator into a fixed pipeline:

The framework samples parents; the model produces one candidate from them. AVO replaces that whole decomposition with a single autonomous agent run:

The agent sees the full lineage Pt, a domain knowledge base K, and the scoring function f — and decides for itself what to read, what to change, and when to measure. It stops being a candidate generator and becomes the variation operator.

This repo implements that framework, plus the surrounding machinery the paper describes: a git-backed lineage, a correctness-gated score vector, the matches-or-improves commit policy, a supervisor that intervenes on stagnation, and trajectory plots. Two optimisation targets ship with it.

The part that matters: it runs on the session you already have

The default driver does not spawn an agent and does not call an API. It hands the variation prompt to the Claude Code session you are already talking to, and that session does the work. Nothing extra is billed, no ANTHROPICAPIKEY is needed, and the agent doing the optimising is a real general-purpose coding agent — which is exactly what the paper used.

Unattended mode (spawn an agent per step and let it run for days, like the paper's 7-day experiment) is available too, and is opt-in precisely because it spends quota.

From the project README.

Maintaining this server?

Add the radar badge to your README — it shows your project was picked up by MCP Radar and links to this page:

[![On MCP Radar](https://mcp.liqiwa.com/badge.svg)](https://mcp.liqiwa.com/s/gatordevin--avo.html)

Related MCP servers

vshulcz/deja-vu

Your agents already solved this. deja finds it — retroactive search and MCP recall over the session logs your coding agents already wrote. Cross-agent, SSH-portable, one zero-dep binary, fully local.

⭐ 391Go
Riccardo8888/agent-link

An end-to-end encrypted channel between coding agents, anywhere

⭐ 55Python
2672243194/agentbrain

Local-first long-term memory for AI agents - plain Markdown vault + thin MCP server. CJK-aware BM25, token-efficient by design.

⭐ 5Python
omaekumiko2-create/kru

Local-first MCP password and credential manager for AI agents. Use passwords, API keys, SSH identities, and TOTP without exposing hidden plaintext to the model.

⭐ 88Rust
pinoox/neuromesh

The Biomimetic Context Engine & Neural Runtime for AI Coding Assistants

⭐ 72Rust
salatmaster/keenetic-mcp

Keenetic router control for Claude Code, Codex, Cursor and any MCP agent. An MCP server over the router's own RCI API, plus ready-made plugins whose skills teach the agent how the hardware really beha

⭐ 18TypeScript

📬 Get the weekly radar in your inbox

The top new MCP servers of the week, every Monday. No spam, unsubscribe anytime.