Token导航 LogoToken导航

Agent Skill · 研究检索

manim-personal-style(马尼姆个人风格)

manim-personal-style 用于查找、检索和筛选相关信息,适合在 Codex、Claude、Cursor、Gemini CLI 中需要根据关键词、任务场景或来源线索快速定位候选结果时使用。可结合来源仓库、安装命令和原始 README 继续核验具体用法。安装前建议确认权限范围、维护状态,以及是否会触发联网、命令执行或文件读写。

manim-personal-style 是研究检索类 Agent Skill,可用宿主为 Codex、Claude、Cursor、Gemini CLI,通过GitHub 仓库获取,权限为只读;许可证为 MIT,第三方扫描已通过,按安装量在研究检索类的 73,675 个 Skill 里排第 66,724。

总安装 259周安装 11GitHub Stars 2下载 91核验 2026-05-01

安装说明

本站只整理中文说明,不托管安装包,也不代用户安装。
GitHub
许可证
MIT

安装方式

通过对话安装

复制提示词发给支持本地命令或 Skills 的 AI 助手,先确认命令和权限,再让它执行。

请帮我安装这个 Agent Skill:manim-personal-style(马尼姆个人风格)
来源仓库:https://github.com/ermaotie/manim-personal-style
仓库路径:skills/manim-personal-style
安装命令:
npx skills add https://github.com/ermaotie/manim-personal-style --skill manim-personal-style
安装前请先检查当前环境是否支持对应 CLI,并向我确认将要执行的命令、安装目录、联网范围和文件读写权限;确认后再执行。

命令行安装

复制命令到本机终端执行。不同来源提供的安装方式可能略有差异;本站展示可直接复制的安装命令,安装前请核对来源页面。

skills.shnpx skills
npx skills add https://github.com/ermaotie/manim-personal-style --skill manim-personal-style

这个 Skill 是做什么的

研究检索类 Skill 主要解决「找资料、核对事实、把来源整理清楚」的问题,通常会让助手先检索再总结。

用于查找、检索和筛选相关信息。适合在 Codex、Claude、Cursor、Gemini CLI 中根据关键词快速定位结果。通过 npx 命令从指定仓库安装并使用。需确认权限范围和维护状态,避免触发联网或文件操作。manim-personal-style 属于研究检索类 Skill,可作为该场景下的辅助能力补充。

来源仓库内路径:skills/manim-personal-style

上手步骤

4 步
  1. 1

    确认自己的宿主工具

    公开资料标注支持 Codex、Claude、Cursor、Gemini CLI;先确认自己用的是其中之一,再按对应工具的 Skills 目录放置。

  2. 2

    拿到安装命令

    可以直接复制的命令是 npx skills add https://github.com/ermaotie/manim-personal-style --skill manim-pe…;命令里的占位符换成自己的路径或密钥再执行。

  3. 3

    安装后重启宿主

    命令执行完要完全退出宿主工具再打开,让新装的 Skill 被重新扫描到,否则列表里可能看不到。

  4. 4

    按最小权限试用

    它标注为只读,先用只读任务试一次,确认行为符合预期再扩大使用范围。

常见问题

6 条

manim-personal-style 怎么安装?

在终端执行 npx skills add https://github.com/ermaotie/manim-personal-style --skill manim-personal-sty…。装完完全退出宿主工具再打开,让它重新扫描本地 Skill。

manim-personal-style 支持哪些 AI 工具?

公开资料标注支持 Codex、Claude、Cursor、Gemini CLI。其他工具只要能读取本地 Skill 目录,一般也能手动放进去,但要自己确认触发方式是否一致。

manim-personal-style 需要什么权限?

权限级别是只读。该 Skill 主要提供规则、说明或参考内容,本身偏只读;真正读写文件、联网或执行命令仍取决于宿主 Agent 的任务。

manim-personal-style 是免费的吗?

来源标注的许可证是 MIT。本站不代收任何费用,也不托管安装包;能否商用、要不要署名,以许可证原文为准。

manim-personal-style 用的人多吗?

公开数据是累计安装 259 次,近一周安装 11 次,下载 91 次,GitHub 2 星。这些数字来自第三方目录统计,只能当作热度参考,不代表质量或安全性。

manim-personal-style 安全吗?

第三方扫描结果为通过,但这只覆盖已知的静态检查项,不能替代自己读一遍源码和安装命令。

值得选的地方和要先想清楚的地方

值得选的地方

  • 同时适配 4 种工具,换客户端不用重新找替代品。
  • 3 个目录同时收录,说明这个 Skill 流传比较广。
  • 许可证明确为 MIT,商用和分发边界好判断。
  • 第三方扫描已通过,常见的高风险行为没有命中。

先想清楚的地方

  • 第三方目录整理,版本更新和行为细节以来源仓库为准。

适不适合你

更适合

  • 已经在用 Codex,想直接扩展它能力的人。
  • 需要处理研究检索类任务、又不想每次都手写提示词的人。
  • 需要许可证清楚、并且有第三方扫描结果的团队项目。

不太适合

  • 需要厂商级技术支持和 SLA 的生产系统。

研究检索类 Skill 对比

6 个同类
Skill总安装GitHub 星权限许可证扫描
find-skills

寻找技能 · Codex、Claude

31,199,81316,486操作浏览器MIT有提醒
self-improving-agent

自我改进 Agent · OpenClaw

16,464,0023,396接触敏感数据MIT-0通过
soultrace

灵魂追踪 · Codex、Claude

7,588,94412需要联网未明确有提醒
self-improving

自我完善 · OpenClaw

6,829,2841,046需要联网MIT-0通过
microsoft-foundry

微软代工厂 · Codex、Claude

6,614,144764调用外部服务未明确有提醒
manim-personal-style

马尼姆个人风格 · 本页

2592只读MIT通过

按安装量取同分类前几名;数字来自第三方目录统计,只作热度参考,不代表质量结论。

SKILL.md 原文

Manim Personal Style

Combine scene composition (narrative-first planning) with ManimCE implementation best practices.

Workflow

Phase 0: Understand the request

  • Ask targeted questions to capture audience, scope, length, style, assets, math depth, visual metaphors, narration, delivery format (mp4/gif), resolution/quality, and constraints (timeline, render limits).
  • If the topic is specialized or likely outdated, do quick research before planning.
  • Confirm any required brand or color palette.

Phase 1: Draft scenes.md (required before code)

  • Always create scenes.md first (use templates/scenes-template.md).
  • Break the video into clear scenes with purpose, visuals, narration notes, and technical notes.
  • Include transitions, recurring motifs, and a color palette.
  • If the user wants to skip planning, still provide a minimal scenes.md and ask for approval.
  • After generating scenes.md, provide a concise overall summary plus per-scene highlights so the user can adjust.

Phase 1.5: Draft scene_with_position.md (required before code)

  • After scenes.md, create scene_with_position.md (use templates/scene-with-position-template.md).
  • For every scene, enumerate all objects, their sizes and positions in frame coordinates.
  • Perform a layout validation: confirm every object stays within the frame bounds and does not overlap others unless explicitly intended.
  • If bounds or overlaps fail, adjust positions/sizes and document the changes.
  • Ask for approval on the layout plan before coding.

Phase 2: Review and lock the plan

  • Ask the user to approve or edit the plan.
  • Validate the plan against user requirements (length, style, audience, delivery format, constraints) and explicitly confirm fit or list mismatches for correction.
  • Only proceed to code after explicit approval or clear confirmation to continue.

Phase 3: Implement with ManimCE best practices

  • Translate each scene into ManimCE code using from manim import *.
  • Prefer one Python file per scene or grouped logically; keep scene names aligned to the plan.
  • Use templates in templates/ to start each scene.
  • Follow the rules in rules/ for animations, timing, text/LaTeX, camera, and styling.
  • Reuse and transform mobjects for continuity; avoid abrupt replacements.
  • If rendering produces multiple video segments, merge them into a single final video without asking.

scenes.md Requirements

  • Must be generated before any code.
  • Must include: overview, narrative arc, scene list with durations, visual elements, narration notes, technical notes, transitions, palette, math content, and implementation order.
  • Save at scenes.md in the working directory unless the user specifies another path.

scene_with_position.md Requirements

  • Must be generated after scenes.md and before any code.
  • Must include for each scene: object list, sizes, positions, anchors, z-order (if relevant), and layout checks.
  • Must include explicit boundary checks (frame width/height or safe margins) and overlap checks.
  • If any conflicts exist, record the adjustments made to resolve them.
  • Save at scene_with_position.md in the working directory unless the user specifies another path.

References (use as needed)

Composition (3b1b-style planning)

  • references/narrative-patterns.md - Narrative arcs and hooks
  • references/visual-techniques.md - Visualization techniques
  • references/scene-examples.md - Example scenes.md fragments
  • templates/scenes-template.md - scenes.md template
  • templates/scene-with-position-template.md - scene_with_position.md template

ManimCE Best Practices

  • rules/scenes.md, rules/animations.md, rules/mobjects.md - core usage
  • rules/text.md, rules/latex.md, rules/text-animations.md - typography/math
  • rules/positioning.md, rules/grouping.md - layout
  • rules/axes.md, rules/graphing.md, rules/3d.md - plotting and 3D
  • rules/camera.md, rules/timing.md, rules/updaters.md - motion control
  • rules/styling.md, rules/colors.md, rules/shapes.md, rules/lines.md - look & feel
  • rules/cli.md, rules/config.md - rendering & config

Examples & Templates

  • examples/ - working ManimCE examples
  • templates/basic_scene.py - standard scene
  • templates/camera_scene.py - MovingCameraScene
  • templates/threed_scene.py - ThreeDScene

Output Expectations

  • Start with scenes.md and request approval.
  • After scenes.md, create scene_with_position.md and request approval.
  • After approval, provide ManimCE implementation with clear scene mapping.
  • After scenes.md, include a short summary + scene-by-scene detail, and invite adjustments.
  • Validate the total video plan and each scene against user requirements (e.g., duration, style, assets) before coding.
  • If output is split across multiple video files, merge into one final video for delivery.
  • Keep code modular and readable; prefer constants for colors and timing.

平台分布

Claude Code

27.18%
按下载量换算25

Codex

21.47%
按下载量换算20

windsurf

18.08%
按下载量换算16

trae

11.76%
按下载量换算11

qoder

7.11%
按下载量换算6

OpenCode

3.42%
按下载量换算3

安全审计

Gen Agent Trust Hub

通过

Socket

通过

Snyk

通过

权限和风险

只读

该 Skill 主要提供规则、说明或参考内容,本身偏只读;真正读写文件、联网或执行命令仍取决于宿主 Agent 的任务。

安装前确认

本站仅展示第三方公开信息,不托管安装包,不提供自动安装或运行环境。安装前应自行审查源码、依赖和命令行为。

继续浏览同类 Skills