Menu
The review-first prompt is selected by default. You can switch to a direct command or download a local copy.
Read SKILL.md and any companion files shown by SkillsMP before deciding whether to install.
Install with Codex or Claude Copy this prompt, paste it into Codex, Claude, or another assistant, and let it review the skill page and install it for you.
A direct command skips the review prompt. Inspect the source before running it.
npx skills add https://github.com/gszyboy/opencode-config --skill keep-focusedThe command stays on one line. Scroll horizontally to inspect it before copying.
Prefer a local copy? Download the files currently available to SkillsMP.
MiniMax video analysis skill for local mp4/mov/mkv/avi files, especially software operation recordings, 小程序/APP/Web 操作录屏, UX review, interaction analysis, 动效复刻, video analysis, 视频分析, MiniMax-M3, or 本地视频. This skill only analyzes existing video content and does not generate video.
Base-layer skill for the SenseNova-Skills project, providing low-level APIs for image generation, recognition (VLM), and text optimization (LLM). This skill does not preprocess inputs; it only calls backend services and returns results. This skill is not user-facing and is intended for upper-layer skills only.
Creative-mode PPT pipeline. One full-page 16:9 PNG per slide. LLM / VLM calls go through sn-ppt-standard/lib/model_client.py (shared thin client). Text-to-image (the actual png rendering) goes through sn-image-base/scripts/sn_agent_runner.py. Falls back to web image search when T2I generation fails. Expects task_pack.json + info_pack.json already written by sn-ppt-entry.
Based on SOC occupation classification
Showing SKILL.md
| name | keep-focused |
| description | 防止 AI 开发过程中跑偏、越改越多、过度工程。当 AI 开始"顺便..."、"优化了..."、"重构了..."时立即触发。 |
以下任一情况出现时，立即执行"拉回机制"：
| 信号 | 示例 | 处理 |
|---|---|---|
| AI 说"顺便..." | "我顺便优化了..." | /undo + 拒绝 |
| AI 说"我重构了..." | "我把整个模块重构了" | /undo + 拒绝 |
| AI 说"为了更好的扩展性" | "添加了一个抽象层" | /undo + 拒绝 |
| 超出 Change Budget | 修改了 4+ 文件 | /undo + 重新约束 |
| 涉及任务外文件 | 修改了无关模块 | 拒绝 + 重新明确范围 |
| 信号 | 处理 |
|---|---|
| AI 开始添加新文件 | 确认是否必要 |
| AI 提到设计模式 | 提醒当前是中小型项目 |
| AI 提到 service/repository | 提醒规范要求 |
| AI 输出超过 200 行 | 提醒精简 |
检测到以下信号：
- [具体信号]
- 任务范围：[原定范围]
- 实际改动：[超出部分]
# 如果 AI 已经做了改动
/undo
# 或者拒绝
"你的改动超出了任务范围。请：
1. /undo 撤销刚才的改动
2. 只实现我要求的功能
3. 其他需求我会另外开任务"
任务：[原任务描述]
Change Budget：
- 文件数：≤ 3
- 行数：≤ 80
- 不修改：[具体排除的文件/模块]
请继续完成任务。
"对话已经较长，为防止上下文混乱：
1. 先完成当前任务
2. /new 新建会话继续下一个任务
3. 或 /compact 压缩上下文"
"你连续多次试图扩展任务范围。这违反了 Change Budget 原则。
请只完成我最初要求的任务，其他需求另外开任务讨论。"
用户："添加一个按钮"
AI："好的，我来重构一下整个组件结构..."
→ 立即拉回，只添加按钮
用户："修改用户详情页"
AI："我还要修改一下相关的 API..."
→ 拒绝 API 修改，只改页面
用户："实现登录功能"
AI："我来加个记住密码和自动登录..."
→ 拒绝，先完成基本登录
1. 小 ≠ 不完整
2. 简单 ≠ 不专业
3. 够了 ≠ 要更好
4. 当前任务 ≠ 全部任务