Sync ParaSkill amplify-genuine-strengths

LICENSE

@@ -0,0 +1,21 @@
+MIT License
+
+Copyright (c) 2026 bhxdianzhang
+
+Permission is hereby granted, free of charge, to any person obtaining a copy
+of this software and associated documentation files (the "Software"), to deal
+in the Software without restriction, including without limitation the rights
+to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
+copies of the Software, and to permit persons to whom the Software is
+furnished to do so, subject to the following conditions:
+
+The above copyright notice and this permission notice shall be included in all
+copies or substantial portions of the Software.
+
+THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+SOFTWARE.

README.md

@@ -0,0 +1,56 @@
+---
+license: mit
+language:
+- zh
+- en
+tags:
+- codex-skill
+- ai-agent
+- scientific-writing
+- research-workflow
+- paraskill
+- paradoxgpt
+- writer
+pretty_name: AmplifyGenuineStrengths
+---
+
+# AmplifyGenuineStrengths
+
+`amplify-genuine-strengths` is a ParaSkill research workflow skill for the `writer` family.
+
+## Purpose
+
+Make the paper's REAL strengths visible to the reviewer — articulate them clearly and place them where they count. Grounded in ARGAR's finding that highlighting strengths is the stable, effective lever (impress > convince), but applied HONESTLY: only amplify strengths that are backed by evidence in the paper; flag any "strength" with no supporting evidence so it is NOT inflated. Output is a placement/phrasing plan, not a rewrite of evidence.
+
+## Metadata
+
+- Family: `writer`
+- Owner model: `ParadoxGPT-Writer-4B`
+- Output type: `revision`
+- Version: `0.1`
+- Default language: `zh`
+
+## Included Files
+
+```text
+AmplifyGenuineStrengths/
+  SKILL.md
+  agents/openai.yaml
+  README.md
+  LICENSE
+```
+
+`SKILL.md` is the executable skill instruction. `README.md` is the Hugging Face card.
+
+## Installation
+
+Copy this folder into a Codex-discoverable skills directory, or reference `SKILL.md` directly from an agent workflow.
+
+```bash
+mkdir -p ~/.codex/skills
+cp -r AmplifyGenuineStrengths ~/.codex/skills/amplify-genuine-strengths
+```
+
+## Source
+
+This skill is part of the ParaSkill library for ParadoxGPT. The full routing table is maintained in `SKILL_INDEX.json` in the source bundle.

SKILL.md

@@ -0,0 +1,115 @@
+---
+name: amplify-genuine-strengths
+description: >-
+  Make the paper's REAL strengths visible to the reviewer — articulate them
+  clearly and place them where they count. Grounded in ARGAR's finding that
+  highlighting strengths is the stable, effective lever (impress > convince), but
+  applied HONESTLY: only amplify strengths that are backed by evidence in the
+  paper; flag any "strength" with no supporting evidence so it is NOT inflated.
+  Output is a placement/phrasing plan, not a rewrite of evidence.
+metadata:
+  owner_model: ParadoxGPT-Writer-4B
+  skill_family: writer
+  version: "0.1"
+  output_type: revision
+  default_language: zh
+  trigger_keywords: [strength, strengths, highlight, 放大优点, contribution, selling point, impress, 卖点, contribution list]
+  required_inputs: [paper_context, strengths_evidence]
+  optional_inputs: [target_section]
+  handoff_to: [revise-by-criterion, enforce-content-preservation-zones, diagnose-aha-moment]
+---
+
+# Amplify Genuine Strengths
+
+## Purpose
+
+把论文**真实存在、有证据支撑**的优点，清楚地讲出来、放到该放的位置，让审稿人能看到。依据 ARGAR 的 strength-weakness asymmetry：**让 strengths 显眼是稳定有效的杠杆**（86.1% 的轮次有效，远超消 weakness）。但本 skill 只做诚实放大——无证据的"优点"必须标出、不得夸大，否则就是 ARGAR 揭露的 presentation gaming。
+
+## When to Use
+
+- 论文有真东西，但 strengths 被埋没、写得弱、或没放进 contribution list。
+- 想在不加实验的前提下，让审稿人更容易看到价值。
+- 投稿前把卖点显式化。
+
+## Do Not Use
+
+- 优点本身没证据支撑 → 不要放大（应转 `detect-presentation-gaming` 或补实验）。
+- 想通过夸大措辞提升 perceived merit → 禁止（这是 gaming）。
+- 需要改具体段落文本 → 用 `revise-by-criterion` 执行。
+
+## Required Inputs
+
+- `paper_context`: 全文（够定位证据）
+- `strengths_evidence`: 候选优点清单，每条尽量带论文内证据（section/figure/数字）
+
+## optional Inputs
+
+- `target_section`: 想在哪里显眼化（abstract / intro contribution list / discussion）
+
+## Output Contract
+
+输出必须包含：
+
+1. **genuine_strengths** — 每条：优点 + 论文内证据 + 当前是否被埋没。
+2. **placement_plan** — 每条优点该出现在哪、怎么措辞（动词化、具体化，但不夸大）。
+3. **unsupported_flags** — 无证据支撑、**不应**放大的"优点"，标注原因（防 gaming）。
+4. **protected_strengths** — 审稿人已认可的优点，后续修改不得削弱（ARGAR 的 protected 机制）。
+
+## Procedure
+
+1. 逐条核对 strengths_evidence 是否有论文内真实证据。
+2. 有证据的 → 判断是否被埋没 → 给 placement + 措辞建议（具体、可核查，不堆 "novel/comprehensive"）。
+3. 无证据的 → 进 unsupported_flags，明确说不要放大。
+4. 标出应被保护的已有优点。
+
+## Quality Bar
+
+一个好的输出必须：
+
+- 每条 genuine_strength 都指向论文内可核查的证据。
+- 措辞建议是"把已有事实讲清楚"，不是拔高。
+- unsupported_flags 诚实标注，防止 gaming。
+
+一个差的输出：
+
+- 把无证据的点当优点放大。
+- 用 "first/fundamental/comprehensive" 等词拔高。
+- 笼统说"加强贡献"而不指具体证据。
+
+## Failure Modes
+
+- **strength inflation**：把局部结果包装成普遍贡献（= ARGAR 的 strength inflation，应转 `detect-presentation-gaming`）。
+- **parroting self-positioning**：照搬论文自我标定词当优点。
+- **削弱已认可优点**：改写时丢了 reviewer 认可的 strength。
+
+## Handoff
+
+- 要把 placement 落到具体段落文本 → `revise-by-criterion`
+- 担心某条优点其实越界/无依据 → `enforce-content-preservation-zones`
+- 想先诊断 strengths 是否对得上核心 realization → `diagnose-aha-moment`
+
+## Example
+
+Input:
+
+```
+paper_context: <interaction scaling 论文>
+strengths_evidence:
+  - "长程任务曲线不饱和" (exp Fig2)
+  - "方法简单" (method)
+target_section: introduction (contribution list)
+```
+
+Output（节选）:
+
+```
+genuine_strengths:
+  - 优点: interaction scaling 带来不饱和提升。证据: Fig2 三任务曲线。当前: 埋在 exp，intro 未提。
+  - 优点: 机制简单(仅累积)。证据: method §3.2。当前: 未强调。
+placement_plan:
+  - intro contribution list 第1条: "We show interaction rounds are a scalable resource,
+    yielding unsaturated gains on long-horizon tasks (Fig2)." —— 具体到 Fig2，不拔高。
+  - 第2条: "The gain is obtained by a simple accumulation mechanism." —— 如实。
+unsupported_flags: []
+protected_strengths: ["不饱和提升"]  # 后续修改不得削弱
+```

agents/openai.yaml

@@ -0,0 +1,4 @@
+interface:
+  display_name: "Amplify Genuine Strengths"
+  short_description: "Make the paper's REAL strengths visible to the reviewer — articulate them clearly and plac"
+  default_prompt: "Use $amplify-genuine-strengths for this research-writing task."