$npx -y skills add joshua-zyy/academic-paper-writer --skill academic-polishingPolish academic prose, de-AI-ify text, control claim strength, or rewrite method sections for CS/AI/ML papers. Executes Prose Quality Gate, Claim Strength Audit, and de-AI pass. Use when: removing AI writing patterns from paper text, adjusting claim strength to match evidence lev
| 1 | # Academic Polishing |
| 2 | |
| 3 | 将此 skill 视为"学术文体打磨代理"——不是简单润色,而是执行 prose 质量闸门、去AI化改写、claim 强度控制和 Method 专项叙事强化。 |
| 4 | |
| 5 | ## Router Protocol |
| 6 | |
| 7 | 1. Read `manifest.yaml`. It declares `always_load` files, `axes`, and `references.on_demand`. |
| 8 | 2. Read every file listed under `always_load`. These are the skill's binding rules — not reference material. |
| 9 | 3. Apply the loaded material as constraints: |
| 10 | - `stance.md` defines non-negotiable rules, AI Traffic Light, claim strength, evidence-aware boundary, and scope. |
| 11 | - `red-lines.md` defines absolute prohibitions. Do not negotiate these. |
| 12 | - `output-contract.md` defines deliverables and completion criteria per mode. |
| 13 | - `anti-patterns.md` defines known failure modes and their correct alternatives. |
| 14 | 4. Detect the mode using the manifest's `mode` axis. If ambiguous, ask one concise question. |
| 15 | 5. Echo the selected mode to the user before executing. |
| 16 | 6. Reach for `references/` only when the manifest's `references.on_demand` condition is satisfied. |
| 17 | |
| 18 | **Structural debt is not language debt.** If `section_contract_debt = open`, only apply local safe edits and return the diagnosis. |
| 19 | |
| 20 | ## Modes |
| 21 | |
| 22 | | Mode | Use when | |
| 23 | |---|---| |
| 24 | | `prose-quality-gate` | General prose quality check + rewrite | |
| 25 | | `method-prose-rewrite` | Method section narrative: 问题→设计→机制→收益/边界 | |
| 26 | | `de-ai-pass` | Remove AI writing patterns only | |
| 27 | | `claim-strength-audit` | Audit and adjust claim strength | |
| 28 | |
| 29 | ## Execution |
| 30 | |
| 31 | Called internally by `academic-paper-writer` orchestrator at Step 9.6. The main agent reads this file and references/ to execute polishing in-process (no subagent dispatch). Polishing must not modify project source code, configuration, or data files. Max 2 rewrite rounds. |
| 32 | |
| 33 | ## Independent Use |
| 34 | |
| 35 | | Input | Mode | Priority | Behavior | |
| 36 | |---|---|---|---| |
| 37 | | "润色一下", "改改语言" | prose-quality-gate | 2 (fuzzy) | Quality gate + rewrite, max 2 rounds | |
| 38 | | "去掉 AI 味", "不像人类写的" | de-ai-pass | 2 (fuzzy) | AI pattern removal only | |
| 39 | | "claim 太强了", "降级结论" | claim-strength-audit | 2 (fuzzy) | All claims audited, zero-tolerance check | |
| 40 | | "改写 Method", "Method 写得太潦草" | method-prose-rewrite | 2 (fuzzy) | Method narrative: problem→design→mechanism→benefit/boundary | |
| 41 | | Explicit mode name | as specified | 1 (explicit) | Skip inference, use specified mode | |
| 42 | |
| 43 | | Scenario | Recommended | |
| 44 | |---|---| |
| 45 | | Just polish/de-AI text | This skill (standalone) | |
| 46 | | Auto-polish during drafting | academic-paper-writer orchestrator (Step 9.6) | |
| 47 | | After revision review | academic-reviser → this skill | |