cn-patent-disclosure-review
Skill4 starsChanged 3 months agoMITView on GitHub
当用户已有技术版交底书初稿(通常由 `cn-patent-disclosure-draft` 产出),需要引导发明人在终稿决策前逐阶段审核并反馈实质性偏差(组件遗漏、因果链不成立、解决思路工程不可行等)时使用。仅做发明人审核与反馈收集,不重做评审、不改写交底书;累积反馈触发整体修订时由编排器派回 `cn-patent-disclosure-draft` 重写。触发词:发明人审核、交底书审核、inventor review、disclosure review、step-3.inventor-review。
In this repository14 other files it gives its agents.
- cn-patent-attorney-reviewSkill
- cn-patent-disclosure-draftSkill
- cn-patent-docx-exportSkill
- cn-patent-domain-runtimeSkill
- cn-patent-formal-draftingSkill
- cn-patent-mainline-analysisSkill
- cn-patent-prior-art-searchSkill
- cn-patent-repo-scoutSkill
- generate-imageSkill
- markdown-mermaid-writingSkill
- matplotlibSkill
- scientific-schematicsSkill
- scientific-visualizationSkill
- seabornSkill
The file
Files
发明人交底初稿审核 skill
主跑型 skill,在 step-3.inventor-review 阶段由 cn-patent-domain-runtime 编排器加载,引导用户对交底初稿做 3 阶段审核。
入口前置
state.current_stage == "step-3.inventor-review"patent/<patent-slug>/disclosure/disclosure-draft.md已落盘state.step_3.disclosure_draft.status == "completed"
由 scripts/validate-stage.py --stage step-3.inventor-review --mode enter 自动校验。
行为前置
- 编排器禁止把 3 段 AskUserQuestion 合并成综合题;每段按
references/protocols.md独立调用,提示词与写入字段不可省略 - 加载后第一动作:把当前 ISO 时间写入
state.step_3.inventor_review.skill_loaded_at(exit 校验依赖该字段)
3 阶段审核流程
| 阶段 | 看交底书章节 | AskUserQuestion 三档 |
| Stage 1 组件拆解核对 | §4.1 + §4.2-§4.X | 无遗漏 / 需补充修正 / 整段重写 |
| Stage 2 因果链与创新点合理性 | §3 + §4.X 五栏表 + §5 | 合理成立 / 部分调整 / 关键站不住 |
| Stage 3 解决思路可行性 | §4 + §6.1 + §6.2 | 全可行 / 部分不可行 / 主方案不可行 |
详细提示词、选项文案、状态字段写入见 references/protocols.md。
反馈处理
- Stage 任一档不是「通过」 → 反馈累积进
state.step_3.inventor_review.stage_X_feedback,3 阶段全跑完再一次性整体修订 - 修订时调
cn-patent-disclosure-draft重生成 disclosure-draft.md,复用其起草纪律 - 硬限 3 轮 → 触发 escalation 三档(见 protocols.md)
REQUIRED SUB-SKILL: 修订调用 cn-patent-disclosure-draft 时复用其 references/drafting-discipline.md 作准绳。
REQUIRED CONTEXT: 状态字段约定见 cn-patent-domain-runtime/references/state-machine-reference.md。
出口分支
写入 state.step_3.inventor_review.exit_status + gate_passed,由编排器读取后路由:
| exit_status | gate_passed | 编排器路由 |
approved | true | step-3.post-disclosure-decision |
accepted_with_dissent | true | step-3.post-disclosure-decision(未解决项由 handoff S1 读 stage_X_feedback 展示) |
rolled_back_step_1 | false | current_stage = step-1,reset step_2/step_3 下游字段 |
rolled_back_step_2 | false | current_stage = step-2,reset step_3 下游字段 |
跨平台 fallback
主跑型 skill,不派 subagent,所有交互通过主 agent 的 AskUserQuestion 完成;Claude Code / Cursor / Codex 行为一致。
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

