claims-evidence

スター1

フォーク0

更新日2026年3月16日 19:56

Map claims in the paper to results, configs, or code; avoid unsupported claims.

インストール

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。

Manusで実行

ソース

Xmaster6y

Xmaster6y/chess-experiments

GitHub リポジトリを開く Creator のリポジトリを見る

ダウンロード

Manusで実行

Claims–Evidence Skill

When to Use

You are drafting or revising paper text and need to ensure every claim is backed by evidence.
You are reviewing the paper for unsupported or fabricated claims.

Instructions

Use references/check-rules.md for the mapping rules.
When writing: Before adding a claim (e.g. "our method improves over X"), identify the evidence: result file, config, table, or citation. If none exists, do not add the claim or mark it as TODO.
When reviewing: For each substantive claim in the draft, list the corresponding evidence (path or citation). Flag any claim with no evidence.
Never fabricate: Do not invent numbers, tables, or figures. Ask the user for real outputs or suggest running the experiment first.

このリポジトリの他の Skills

同じリポジトリ

experiment-implementation

Xmaster6y/chess-experiments

How to change code, tests, configs, and launch in this repo; output is concrete edits and run commands.

2026-03-171

lczerolens

Xmaster6y/chess-experiments

Provides guidance for interpretability of Leela Chess Zero (lc0) neural networks. Use when loading lc0 models, encoding chess boards, running inference, visualizing heatmaps, probing concepts, evaluating on puzzles, or pairing with tdhook, captum, zennit, or nnsight for XAI.

2026-03-171

tdhook

Xmaster6y/chess-experiments

Provides guidance for interpreting and manipulating neural network internals using tdhook with TensorDict and PyTorch hooks. Use when needing attribution maps, activation analysis, probing, steering, activation patching, or weight-level interventions on PyTorch or TensorDict models.

2026-03-171

background-writing

Xmaster6y/chess-experiments

Turn grouped literature notes into background or related-work text; no claims beyond what the notes support.

2026-03-161

experimental-design

Xmaster6y/chess-experiments

Patterns and checklists for designing and refining experiments in this research template.

2026-03-161

literature-notes

Xmaster6y/chess-experiments

One-paper-per-note structure and grouping by theme; output is structured notes usable for background/related work.

2026-03-161

name	claims-evidence
description	Map claims in the paper to results, configs, or code; avoid unsupported claims.

Claims–Evidence Skill

When to Use

You are drafting or revising paper text and need to ensure every claim is backed by evidence.
You are reviewing the paper for unsupported or fabricated claims.

Instructions

Use references/check-rules.md for the mapping rules.
When writing: Before adding a claim (e.g. "our method improves over X"), identify the evidence: result file, config, table, or citation. If none exists, do not add the claim or mark it as TODO.
When reviewing: For each substantive claim in the draft, list the corresponding evidence (path or citation). Flag any claim with no evidence.
Never fabricate: Do not invent numbers, tables, or figures. Ask the user for real outputs or suggest running the experiment first.