by KimYx0207 · Claude Skill · ★ 100
Last updated: · Indexed by AgentSkillsHub · Auto-synced every 8h
中文 | English HookPrompt 把随口说出的需求，自动整理成可执行、可验收的专业提示词 因合并关系，本仓库已迁入 Kim Service 合集，不再单独发布新版本。请前往 Kim Service 获取新版本、安装说明和更新记录。 简介 HookPrompt 是一个面向 Claude Code、Codex 等 Agent 工具的 Hook。它会在用户输入进入模型前，把自然语言请求整理成更适合执行的结构化任务简报。 它重点解决三类问题： 用户只说了模糊需求，Agent 不知道真实目标和验收标准。 用户反馈很短，例如“这个不行”“报错了”“帮我看看”，但其实需要诊断或修复。 任务需要清晰边界、成功标准和验证计划，否则容易变成“看起来执行了，但没有真的完成”。 HookPrompt 默认要求模型在首条回复中展示三段式结果： 优化后的完整提示词会使用 role-first、outcome-contract、tagged structure、success criteria 和 verification plan，让后续执行更清楚、更可控。 工作方式 mermaid flowchart LR A["用户输入"] -- B["UserPromptSubmit Hook"] B -- C{"是否需要优化？"} C -- D["返回空结果，原输入继续"] C -- E["生成结构化任务简报"] E -- F["通过 additionalContext 注入模型上下文"] F -- G["A
| Stars | 100 |
| Forks | 28 |
| Language | JavaScript |
| Category | Claude Skill |
| License | MIT |
| Quality Score | 63.2165908660454/100 |
| Last Updated | 2026-08-12 |
| Created | 2025-12-08 |
| Platforms | claude-code, node |
| Est. Tokens | ~14k |
Looking for a HookPrompt alternative? If you're comparing HookPrompt with other claude skill tools, these 1 projects are the closest alternatives on Agent Skills Hub — ranked by topic overlap, star count, and community traction.
Report issues related to MCP integration with Claude here.
Explore other popular claude skill tools:
HookPrompt is Claude Code自动提示词优化Hook：谷歌提示词圣经 + 5任务元提示词 | Auto prompt optimization Hook for Claude Code. It is categorized as a Claude Skill with 100 GitHub stars.
HookPrompt is primarily written in JavaScript.
You can find installation instructions and usage details in the HookPrompt GitHub repository at github.com/KimYx0207/HookPrompt. The project has 100 stars and 28 forks, indicating an active community.
HookPrompt is released under the MIT license, making it free to use and modify according to the license terms.
The top alternatives to HookPrompt on Agent Skills Hub include claude-ai-mcp. Each offers a different approach to the same problem space — compare them side-by-side by stars, quality score, and community activity.
Grades come from a rule-based scan built on the SlowMist agent-security taxonomy, covering 11 red-flag categories including credential harvesting, data exfiltration, and curl | sh installers. It is a first-layer scan, not a manual audit — we say so rather than overstate it.
The scale of the problem is documented independently: Liu et al. (2026), in a study of 31,132 agent skills, report that 26.1% contain security vulnerabilities. Our own full-catalog census is published as a citable open dataset.
Sources & who's responsible: