---
title: "autoopt"
description: "Human-gated, repository-grounded automation for optimization research."
canonical: https://agentpluginsdirectory.com/plugins/autoopt
last-updated: 2026-09-25
---

# autoopt
Human-gated, repository-grounded automation for optimization research.
- Slug: autoopt
- Publisher: Heechang Kim, Ernest Ryu, and Shuvomoy Das Gupta
- Repository: https://github.com/Shuvomoy/AutoOPT
- Manifest: plugins/autoopt/plugin.json
- Version: 0.1.0
- License: Apache-2.0
- Category (editorial): research
- Skills: 8 (auto-opt, bnb-pep-skill, chatgpt-pro-handoff, chatgpt-pro-session, frontier-llm-consult, lean-verify, research-repo-manager, solve-with-highest-reasoning)
- MCP servers: 0
- Stars: 18
- Repository created: 2026-08-03
- Repository last pushed: 2026-08-28
- Publisher type: User
- Listing: https://agentpluginsdirectory.com/plugins/autoopt
- Schema: https://agent-plugins.org/schemas/1.0.0/plugin.schema.json

## What autoopt does, in the publisher's words

AutoOPT is a human-gated, repository-grounded workflow for optimization research. It coordinates numerical algorithm design through performance estimation, approved symbolic reasoning, and Lean verification while preserving researcher control over every consequential transition.

Installing this plugin once makes the complete eight-skill AutoOPT workflow available together. Installation does not provision external runtimes, commercial solvers, browser sessions, accounts, authentication, or host capabilities. Those prerequisites remain separate and are checked by the relevant skill before use.

- auto-opt: orchestrates the human-gated AutoOPT pipeline.
- research-repo-manager: grounds work in a versioned research repository and maintains evidence.
- bnb-pep-skill: formalizes and implements Branch-and-Bound Performance Estimation Problems.
- frontier-llm-consult: routes approved symbolic-discovery and proof-planning consultations.
- chatgpt-pro-session: maintains an attended, reusable ChatGPT Pro research session.
- chatgpt-pro-handoff: packages, previews, submits, monitors, and imports an attended consultation.

From the project README, punctuation lightly normalized. Full text: https://raw.githubusercontent.com/Shuvomoy/AutoOPT/HEAD/plugins/autoopt/README.md

## Skills

- auto-opt: Run, resume, or coordinate a human-gated AutoOPT pipeline session for optimization-research automation inside a versioned research repository. Stages delegate to existing skills by name; Stage 0 repository grounding via research-repo-manager, Stage 1 numerical algorithm design via bnb-pep-skill, St…
- bnb-pep-skill: Formalize, derive, and locally implement Branch-and-Bound Performance Estimation Programming (BnB-PEP) instances from math or plain English. Use when a researcher or agent needs worst-case performance analysis or stepsize/parameter optimization of fixed-step first-order methods via performance esti…
- chatgpt-pro-handoff: Package a full-workspace or directed file-selection handoff for attended ChatGPT Pro consultation through the Chrome plugin at chatgpt.com using GPT-5.6 Sol with Intelligence set to Pro, monitor by waiting in internal chunks of at most 60 seconds while inspecting the ChatGPT UI only roughly every 5…
- chatgpt-pro-session: Persistent ChatGPT Pro workflow for starting, claiming, and continuing one reusable ChatGPT Web conversation with an initial project context bundle, monitored by waiting in internal chunks of at most 60 seconds while inspecting the ChatGPT UI only roughly every 5 minutes after reasoning or generati…
- frontier-llm-consult: Human-gated frontier-model research router for AutoOPT Stage 2 symbolic fitting, candidate formula/proof discovery, proof planning, and independent model review. Use when Codex must choose between an external chatgpt-pro-session consultation and an explicitly invoked native solve-with-highest-reaso…
- lean-verify: Formalize and verify optimization theorem, proof, and certificate targets in Lean/Lake using ChatGPT Pro blueprints only as proof-planning aids and local agentic Lean implementation as the execution path. Use for AutoOPT Stage 3, or standalone optimization-theorem verification, when Codex needs to…
- research-repo-manager: Initialize, refresh, compact, and audit applied-math/optimization research repositories. Use when managing immutable sources, research goals, claim/evidence ledgers, experiment/reproducibility logs, source drift, next sessions, or approval-gated archiving.
- solve-with-highest-reasoning: Run an explicitly invoked, long-horizon native-Codex campaign on one exceptionally difficult mathematical or research problem using a user-confirmed minimum duration, the strongest currently available Codex GPT model, that model's highest supported reasoning setting, adaptive independent subagents,…

Descriptions come from the frontmatter of each SKILL.md, punctuation lightly normalized.