sibyl-heavy
01Sibyl-Research-Team/AutoResearch-SibylSystem
Agent Claude Code
Sibyl Research System heavy-reasoning agent. Used for deep analysis tasks: synthesis, supervision, editing, critical review, and reflection.
Sibyl-Research-Team/AutoResearch-SibylSystem
Agent Claude Code
Sibyl Research System heavy-reasoning agent. Used for deep analysis tasks: synthesis, supervision, editing, critical review, and reflection.
Sibyl-Research-Team/AutoResearch-SibylSystem
Agent Claude Code
Sibyl Research System lightweight agent. Used for quick evaluation tasks: debate roles (optimist, skeptic, strategist), section critique, and cross-critique.
Sibyl-Research-Team/AutoResearch-SibylSystem
Agent Claude Code
Sibyl Research System standard agent. Used for literature research, planning, experiment design, and idea generation.
Sibyl-Research-Team/AutoResearch-SibylSystem
Skill Claude CodeCodex
A skill that asks OpenAI Codex to review work as an independent third party. It adds a separate AI perspective to an existing review.
Sibyl-Research-Team/AutoResearch-SibylSystem
Skill Claude CodeCodex
A writing agent that uses OpenAI Codex, an AI coding and reasoning assistant, to write academic papers. The description does not specify which paper sections or formats it supports.
Sibyl-Research-Team/AutoResearch-SibylSystem
Skill Claude CodeCodex
A research assistant that compares your work with leading results and similar studies. SOTA means the best currently known results in a field.
Sibyl-Research-Team/AutoResearch-SibylSystem
Skill Claude CodeCodex
A research assistant that challenges common assumptions and looks for opposing evidence and overlooked weaknesses.
Sibyl-Research-Team/AutoResearch-SibylSystem
Skill Claude CodeCodex
A research critic that gives strict but fair academic criticism.
Sibyl-Research-Team/AutoResearch-SibylSystem
Skill Claude CodeCodex
A research editor that combines separate paper sections into one complete manuscript.
Sibyl-Research-Team/AutoResearch-SibylSystem
Skill Claude CodeCodex
A Sibyl agent focused on reproducible work, reliable data, and careful experiment design. The description is in Chinese and gives no further details about its specific workflow.
Sibyl-Research-Team/AutoResearch-SibylSystem
Skill Claude CodeCodex
A background experiment supervisor that checks GPU availability, starts queued experiments, and handles changes in runtime or status.
Sibyl-Research-Team/AutoResearch-SibylSystem
Skill Claude CodeCodex
An experiment assistant that writes code and runs research experiments on remote GPUs, which are computers designed for large parallel calculations.
Sibyl-Research-Team/AutoResearch-SibylSystem
Skill Claude CodeCodex
An agent for a final review of research papers at the standard expected by conferences such as NeurIPS and ICML.
Sibyl-Research-Team/AutoResearch-SibylSystem
Skill Claude CodeCodex
A research assistant that decides whether a small experiment should move forward, be revised, or be replaced.
Sibyl-Research-Team/AutoResearch-SibylSystem
Skill Claude CodeCodex
A research assistant that develops bold research proposals by combining ideas from different fields.
Sibyl-Research-Team/AutoResearch-SibylSystem
Skill Claude CodeCodex
A research assistant that brings ideas and methods from fields such as cognitive science, physics, and biology into your work.
Sibyl-Research-Team/AutoResearch-SibylSystem
Skill Claude CodeCodex
A research-data assistant that synchronises data with Feishu cloud storage, an online workspace used to store and share files and information.
Sibyl-Research-Team/AutoResearch-SibylSystem
Skill Claude CodeCodex
A paper-formatting assistant that converts papers to NeurIPS LaTeX format and compiles them. LaTeX is a text-based system for producing structured documents, and NeurIPS is a major machine-learning research conference.
Sibyl-Research-Team/AutoResearch-SibylSystem
Skill Claude CodeCodex
A literature-research assistant that searches arXiv, a public repository for research papers, and the wider web. It supports systematic reviews of existing research.
Sibyl-Research-Team/AutoResearch-SibylSystem
Skill Claude CodeCodex
A method-reviewing agent based on the Sibyl methodology. It examines whether experiments are valid, repeatable, and likely to apply beyond the test itself.
Sibyl-Research-Team/AutoResearch-SibylSystem
Skill Claude CodeCodex
A literature-search assistant that checks whether a research idea is already covered by published papers.
Sibyl-Research-Team/AutoResearch-SibylSystem
Skill Claude CodeCodex
A research assistant that reviews experiment results from a positive angle.
Sibyl-Research-Team/AutoResearch-SibylSystem
Skill Claude CodeCodex
A paper-outlining assistant that creates detailed structures for research papers.
Sibyl-Research-Team/AutoResearch-SibylSystem
Skill Claude CodeCodex
An experiment-planning assistant that designs careful, repeatable experimental procedures. Reproducible means that another researcher can follow the same steps and obtain comparable results.