AGENT SKILL

research-motivation-story-auditor

Use for feasibility-gated research motivation strength auditing, scientific storytelling, reviewer persuasion tests and evidence-backed narrative revision for any user-selected research field. It must not fabricate importance, experiments or novelty.

License 许可
MIT
最近更新
2026-10-13
Stars
★ 0
Forks
0

01 前言 Front Matter

name
research-motivation-story-auditor
description
Use for feasibility-gated research motivation strength auditing, scientific storytelling, reviewer persuasion tests and evidence-backed narrative revision for any user-selected research field. It must not fabricate importance, experiments or novelty.

03 SKILL.md 正文

Research Motivation & Story Auditor — 先问值得做吗,再问论证能说服吗

v3.3 Motivation-first priority

First use ../research-idea-discovery/references/motivation-first-gate.md to audit an independent scientific problem (M0), BEFORE Idea Seeds, D0/G0 and M1. Problem scouting is bounded, neutral, and may include a no-new-model pilot. If PROBE_ONLY/REJECT_OR_REFRAME/RESEARCH_MORE, explicitly refuse architecture, loss, and three contributions: offer a specific motivation-evidence action. The older D0/G0→M1→N/L→M2→S sequence applies only to a problem that passed M0. Do not declare the cause proven because a benchmark showed an unexpected result.

先核实 Idea Seed 已经由插件主动找到数据并通过 D0/G0;先审 M1 动机,再让 N 审查近邻,随后专门审 L 机制逻辑和 M2 动机复审,最后进行 S 科学论证;否则只记录探索性问题和阻碍,不包装论文故事。读取主 Skill references/motivation-and-story-protocol.md、references/feasibility-gate.md、references/novelty-protocol.md。

在近邻检查之前实施 M1,在 N/L 之后进行 M2:痛点有证据、影响真实、根因非臆测、竞争解释和最简对手得到公平对待、独有预测可证伪。失败则回退 E0 或转向。

仅对通过 M1/N/L/M2 的候选实施 S:构造完整、有证据的因果论证链,写 30 秒陈述与 Introduction 纲要;开展三轮相互不同的质疑:动机、机制必要性与冷读审稿。每轮要新证据或实质机制修订,禁止凭形容词拔高,且本质上弱的 Idea 应当被淘汰。

输出 motivation-initial / novelty / mechanism-logic / motivation-recheck / story-card / reviewer-objections / claim-evidence-map / verdict。没有外部独立 Agent 则标 ROLE_SIMULATED。结构检查脚本不是评审质量的自动判定器。

04 同仓库更多 Skills