Agent skill

Kimi3manim

by HarleyCoops in HarleyCoops/KimiK3Manim

Turn math and physics concepts into rendered Manim films. An agent skill from HarleyCoops/KimiK3Manim.

MITAuto-check: notesMedia & Creative

Install Kimi3manim

skills CLI
$ npx skills add HarleyCoops/KimiK3Manim --skill kimi3manim -a claude-code

Project install by default; add -g for ~/.claude/skills/.

GitHub CLI
$ gh skill install HarleyCoops/KimiK3Manim kimi3manim --agent claude-code

Project scope by default; add --scope user for a personal install. Needs GitHub CLI 2.90.0 or later (public preview).

Manual copy
$ git clone --depth 1 https://github.com/HarleyCoops/KimiK3Manim.git skills-src && mkdir -p .claude/skills && cp -r skills-src/skills/kimi3manim .claude/skills/kimi3manim && rm -rf skills-src

Use ~/.claude/skills/ instead of .claude/skills for a personal install. The folder must contain SKILL.md.

Claude Code skills documentation · loads skills from .claude/skills/

Facts

Skill name
kimi3manim
GitHub stars
144
Token cost
~1k tokens
SKILL.md length
495 words
Files
1
Skills in repo
1
Repo updated
First seen
Licence
MIT

At a glance

Turn math and physics concepts into rendered Manim films. An agent skill from HarleyCoops/KimiK3Manim.

  • Works in 6 steps: Locate the repo (config kimi3manim.repo,… → Verify auth: either the user ran uv run… → Warn the user that a full run takes… → …
  • Media & Creative work in your project
  • SKILL.md covers When to Use, Quick Reference, Procedure and Pitfalls, plus 1 more section
  • Calls uv, git and ffmpeg; needs MOONSHOT_API_KEY

What it does

Kimi3manim is an agent skill from HarleyCoops/KimiK3Manim. Turn math and physics concepts into rendered Manim films

Its SKILL.md is about 1k tokens, which your agent loads only when the skill is triggered. It is a single SKILL.md file with no bundled scripts.

It sits in Media & Creative. It works with Manim and Kimi. The repository describes itself as: Create epic math and physics explainer animations with Kimi K3. The licence is MIT.

When your agent uses it

  • Media & Creative work in your project

Example prompts

  • “/kimi3manim”

Requirements

  • Python 3
  • A credential in MOONSHOT_API_KEY

Workflow steps

6 steps, taken from the first numbered list in SKILL.md.

  1. Locate the repo (config kimi3manim.repo, default ~/KimiK3Manim).
  2. Verify auth: either the user ran uv run kimi login once (Kimi Code
  3. Warn the user that a full run takes minutes and spends Kimi quota
  4. Phrase the concept the way the intended audience would ask it - the
  5. Run the supervisor command. The built-in preflight aborts with exact
  6. Outputs land in output/k3_runs//: numbered JSON artifacts

What it can do on your machine

Read from SKILL.md and the folder at commit ed954d1. It shows what the files ask for, not the result of running them.

  • Tool permissions

    Pre-approves nothing: there is no allowed-tools line, so your agent's usual permission prompts apply.

    From allowed-tools in the SKILL.md frontmatter.

  • Runs code

    Shell commands in SKILL.md call:

    • uv
    • git
    • ffmpeg

    From the folder's file list and the shell code blocks in SKILL.md.

  • Network

    No URLs in SKILL.md. Its commands use uv and git, which can reach the network depending on how they are called.

    From URLs in SKILL.md, links to its own repository left out.

  • Credentials

    Names these keys or tokens, usually read from environment variables:

    • MOONSHOT_API_KEY

    From names ending in _API_KEY, _TOKEN, _SECRET, _KEY or _PASSWORD in SKILL.md.

Context cost

Kimi3manim loads about 1k tokens when it runs. Until then it costs about 17 tokens; SKILL.md has 495 words of instructions outside code blocks.

Always · name and description, kept in context so the agent knows when to use it
~17
When it runs · the whole SKILL.md, loaded when a task matches
~1k

Estimates: characters ÷ 4, the usual rule of thumb; real counts depend on the model's tokenizer. Scripts and assets cost tokens only if the agent reads them.

Safety

Auto-check: notes

The automated check noted patterns worth knowing about, such as sudo or a known installer.

  • NoteMentions a .env fileSKILL.md:58
    `.env` (also enables the vision critic stage).

Automated static check — not a guarantee. Review scripts before installing. It scans the text of SKILL.md for risky patterns (piping downloads into a shell, reading credential files, hidden Unicode, destructive commands); files beside SKILL.md are not scanned.

SKILL.md

The full file from HarleyCoops/KimiK3Manim at commit ed954d1, republished under its MIT licence (© HarleyCoops). 495 words, ~1,037 tokens.

Download SKILL.mdSave it as .claude/skills/kimi3manim/SKILL.md (or your agent's skills folder).
name
kimi3manim
description
Turn math and physics concepts into rendered Manim films
version
1.0.0
author
HarleyCoops
license
MIT

Kimi3Manim: concept to rendered film

Drive the Kimi3Manim six-agent pipeline (Concept Scout, Mathematical Enricher, Visual Designer, Narrative Composer, Manim Coder, vision Render Critic under a deterministic supervisor) to turn any math or physics concept into a rendered 1080p Manim film.

When to Use

  • The user asks to animate, visualize, or make a video/film of a mathematical or physical concept, at any level from middle-school geometry to graduate physics.
  • The user wants to resume or debug a failed animation run.
  • The user wants an existing Manim scene re-rendered.

Quick Reference

TaskCommand (run in the repo directory)
Full pipeline, concept to mp4uv run python -m k3_agents.supervisor "<concept>" --quality qh
Resume failed run at Stage 5uv run python resume_pipeline.py [run_dir]
Re-render a scene, no model callsuv run manim -qh <file.py> <SceneClass>
One-time subscription loginuv run kimi login
Install depsuv sync

Quality flags: ql 480p (smoke test), qm 720p, qh 1080p, qk 4K. --max-repairs N bounds coder/critic repair rounds (default 3).

Procedure

  1. Locate the repo (config kimi3manim.repo, default ~/KimiK3Manim). Missing? git clone https://github.com/HarleyCoops/KimiK3Manim.git, then uv sync inside it. System deps: ffmpeg, texlive, texlive-latex-extra, dvisvgm.
  2. Verify auth: either the user ran uv run kimi login once (Kimi Code OAuth subscription - the default), or MOONSHOT_API_KEY is set in .env (also enables the vision critic stage).
  3. Warn the user that a full run takes minutes and spends Kimi quota (six kimi-k3 calls at max reasoning plus rendering); never launch runs in parallel.
  4. Phrase the concept the way the intended audience would ask it - the Scout calibrates prerequisite depth from the phrasing. "why the Pythagorean theorem is true" is as valid as "the double cover of SO(3)".
  5. Run the supervisor command. The built-in preflight aborts with exact instructions if the environment is broken - relay its message verbatim and fix the environment; do not retry the pipeline.
  6. Outputs land in output/k3_runs/<slug>/: numbered JSON artifacts (01 knowledge graph ... 06 critique), generated scenes in scenes/, per-attempt render_attemptN.log, and the mp4 under media/. Deliver the mp4. [[as_document]]
Show full SKILL.md (158 more words)Show less

Pitfalls

  • Environment failures (missing manim/latex/pango) are NOT code bugs: the supervisor aborts on them by design. Fix the environment; never feed them to the coder repair loop.
  • Re-running the same concept reuses its run directory - move it aside first for a clean slate.
  • Subscription-only auth cannot run the Stage 6 vision critic; the run stops after rendering with the mp4 already on disk. Ship it and note the critic was skipped, or set MOONSHOT_API_KEY to enable it.
  • On WSL, keep the clone in the Linux filesystem (not /mnt/c) or uv and renders run slowly; export UV_LINK_MODE=copy silences the hardlink warning.

Verification

  • The run directory contains an mp4 under media/ and artifacts 01-05.
  • On failure, read the newest render_attemptN.log in the run dir BEFORE re-running; it contains the real traceback.
  • Sample a few frames (ffmpeg -ss <t> -i <mp4> -frames:v 1 f.png) and confirm equations are legible and scenes match the visual spec before presenting the film as final.

© HarleyCoops, MIT. Rendered from Markdown: HTML in the file is shown as text, images as links, and headings moved down two levels. Raw file

Files

Just SKILL.md in skills/kimi3manim of HarleyCoops/KimiK3Manim.

Open the folder on GitHubat commit ed954d1

Compare with similar skills

Kimi3manim next to the 5 skills that share the most tags, products or categories with it. Stars are the repository's; “used in” counts other GitHub owners with a copy.

Kimi3manim compared with similar skills
SkillStarsUsed inTokensAuto-checkLicenceRepo updated
Kimi3manim this skillHarleyCoops/KimiK3Manim144—~1kAutomated safety check: NotesMIT
Qiaomu Cutjoeseesun/qiaomu-cut-skill372—~6.8kAutomated safety check: NotesMIT
ShowtimeFavioVazquez/showtime220—~3kAutomated safety check: PassMIT
Noon Authoringyongkyuns/noon133—~2kAutomated safety check: PassNone
Animatecth9191/animate319—~4.5kAutomated safety check: PassMIT
Caption Generatorszhshp/panda-video-generator291—~1.5kAutomated safety check: NotesNone

Similar skills

  • Qiaomu Cut

    joeseesun/qiaomu-cut-skill

    把一句话需求转成可复现、可验收视频工程的乔木智能剪辑导演。Use when the user asks to create, plan, edit, remix, explain, narrate, subtitle, animate, composite, or render a video—including one-line requests such as “制作一个科普视频:介绍…

    372 GitHub stars~6.8k tokensUpdated 12 days ago
    Media & CreativeAuto-check: notes
  • Showtime

    FavioVazquez/showtime

    A skill your agent uses when the user wants a video made, edited or finished: a launch or promo, product demo, explainer, trailer or teaser, tutorial or walkthrough, a screen recording turned into a…

    220 GitHub stars~3k tokensUpdated yesterday
    Media & CreativeAuto-check passed
  • Noon Authoring

    yongkyuns/noon

    Author and review Noon animation scenes using capability-qualified ManimCE-style Python or the shared Rust API.

    133 GitHub stars~2k tokensUpdated today
    Media & CreativeAuto-check passed
  • Animate

    cth9191/animate

    Make a short procedural animation in any style — an explainer, a history, a little story — as a single-file canvas video with a synthesized score, rendered to MP4.

    319 GitHub stars~4.5k tokensUpdated 4 days ago
    Media & CreativeAuto-check passed
  • Caption Generator

    szhshp/panda-video-generator

    Prefer the hosting AI Agent to write narration into input.txt from crawl JSON (no external LLM).

    291 GitHub stars~1.5k tokensUpdated 2 mo ago
    Media & CreativeAuto-check: notes
  • Remotion

    zhuzhaoyun/Molio

    Molio's builtin skill for MAKING a video from any source — wiki notes, articles, scripts, product info, or a brief — and rendering it to MP4.

    433 GitHub stars~4k tokensUpdated today
    Media & CreativeAuto-check passed

Works with

Questions about Kimi3manim

What does Kimi3manim do?

Turn math and physics concepts into rendered Manim films. An agent skill from HarleyCoops/KimiK3Manim. Kimi3manim is an agent skill from HarleyCoops/KimiK3Manim.

When should I use Kimi3manim?

Kimi3manim fits situations like: media & Creative work in your project.

How do I install Kimi3manim in Claude Code?

Run `npx skills add HarleyCoops/KimiK3Manim --skill kimi3manim -a claude-code`. Or copy the skill folder (skills/kimi3manim in HarleyCoops/KimiK3Manim) into .claude/skills/kimi3manim in your project. Claude Code loads it when a task matches its description.

How do I install Kimi3manim in Codex?

Run `npx skills add HarleyCoops/KimiK3Manim --skill kimi3manim -a codex`. Or copy the skill folder (skills/kimi3manim in HarleyCoops/KimiK3Manim) into .agents/skills/kimi3manim in your project. Codex loads it when a task matches its description.

Can I use Kimi3manim in Cursor, Gemini CLI or GitHub Copilot?

Cursor, Gemini CLI, GitHub Copilot and OpenCode also load SKILL.md folders. With the skills CLI, run `npx skills add HarleyCoops/KimiK3Manim --skill kimi3manim -a cursor` (or -a gemini-cli, github-copilot or opencode for the others). To copy it by hand, put the folder in .cursor/skills/kimi3manim, .gemini/skills/kimi3manim, .github/skills/kimi3manim and .opencode/skills/kimi3manim in your project.

What does Kimi3manim need to run?

Going by SKILL.md and its folder, Kimi3manim needs the command-line tools its instructions call (uv, git and ffmpeg) and credentials named MOONSHOT_API_KEY. Our summary lists: Python 3; A credential in MOONSHOT_API_KEY.

Does Kimi3manim access the network?

SKILL.md contains no URLs. Its commands use uv and git, which can reach the network depending on how they are called. This is read from the text; nothing was executed.

Is Kimi3manim safe to install?

Our automated static check of SKILL.md found notes only (mentions a .env file), nothing it rates as a warning. It is not a guarantee. Review the folder before installing.

What licence does Kimi3manim use?

Kimi3manim is published under the MIT licence (declared in SKILL.md). It allows redistribution, so the full SKILL.md is shown on this page.

How many tokens does Kimi3manim use?

About 1k tokens (SKILL.md is roughly 4.1k characters). Agents keep only the skill's name and description in context until a task matches; then they load SKILL.md in full.

What are the alternatives to Kimi3manim?

Skills that share tags, products or a category with Kimi3manim: Qiaomu Cut (joeseesun/qiaomu-cut-skill, 372 stars), Showtime (FavioVazquez/showtime, 220 stars), Noon Authoring (yongkyuns/noon, 133 stars) and Animate (cth9191/animate, 319 stars). The comparison table on this page puts their stars, adoption, token cost, safety result and licence side by side.

Who maintains Kimi3manim?

HarleyCoops (a GitHub user) maintains it in HarleyCoops/KimiK3Manim, which has 144 GitHub stars. The repository was last updated on July 29, 2026.

Source: HarleyCoops/KimiK3Manim on GitHub. Facts on this page come from the repository at the commit we read; the author's words are quoted as theirs.