メニュー
Defines the structure and requirements for a research proposal.
Codex または Claude でインストール この Prompt をコピーして Codex、Claude、または他のアシスタントに貼り付けると、Skill ページを確認してインストールできます。
SOC 職業分類に基づく
Defines how to save and load agent state as checkpoint files.
Mechanism for keeping an agent anchored to a caller-defined persona across sessions. Covers the biplate private memory system for drift detection and self-correction.
Multi-agent research war room. Personas debate in sequential turns through two phases — ideation and proposal writing. Persona persistence and drift detection are enforced every turn via the persistent-persona skill.
Search and summarize papers from ArXiv. Use when the user asks for the latest research, specific topics on ArXiv, or a daily summary of AI papers.
Core research principles that govern all project decisions. Load this skill for every research project. Every idea, method, and artifact must be evaluated against these principles before proceeding.
Probe system resources — CPU, RAM, disk, GPU/CUDA/MPS — and translate findings into experiment design recommendations. Run this at the start of any compute-intensive project.
| name | research-proposal |
| type | instruction |
| description | Defines the structure and requirements for a research proposal. |
A research proposal is a structured 2-page document. Every section must be present. Be concrete — vague claims are not acceptable.
What does the system do? What is its input and output? What real-world problem does it solve? Scope must be neither too narrow nor too broad.
Concrete examples of inputs and outputs. Describe what the input and output variables look like and how they interact. Reference the actual data being used — do not describe hypothetical data.
How will success be measured? Why is this metric the right one for this problem? A reasonably sized dataset of example input-output pairs must be identified — either from an existing source or collected from scratch.
What existing research and projects are relevant? How does the proposed work differ? Reference specific papers and systems, not general fields.
How will the problem be approached? Why does this method fit? What are the key challenges? What algorithms or models are appropriate, and what are the tradeoffs between accuracy and efficiency?
