Home / All Skills / Content & Media Creation / after-effects
Content & Media Creation · iart-ai/motion-design-skills

after-effects

This skill should be used when the user asks anything about Adobe After Effects — to "write an After Effects expression", "make a wiggle / loopOut / inertial bounce", "rig a control with a slider / null", "build a MOGRT / Essential Graphics template", "export AE to Lottie or Media Encoder"; OR to "export from After Effects", "export with transparency / an alpha channel", "fix a black background on export", "choose ProRes vs H.264", "which codec / what bitrate for 1080p / 4K", "stop H.264 disappearing from the Render Queue"; OR when "After Effects is slow / laggy / crashing / froze", "RAM preview won't play in real time", "AE is using all my memory / render takes forever", "how do I speed up After Effects"; OR to "organize my After Effects project", "relink missing footage", "fix broken links after moving files", "set a folder structure / naming conventions", "batch rename layers", or "write an ExtendScript / .jsx script". Covers expressions & rigging, export recipes, performance triage, and project hygiene.

L4 High-Privilege · Graded auth & full auditV1 Commit-PinnedExecute CommandWrite FileDual-pass Audit
Author iart-aiGitHub iart-ai/motion-design-skills ↗Stars 35License MIT(仓库根 LICENSE:'MIT License' / Copyright (c) 2026 iart.ai;.claude-plugin/plugin.json 的 license 字段亦为 MIT)Commit 3c129f769d
Control levels describe skill capabilities and suggested control measures; they do not represent quality or safety certifications, nor do they dictate usability. Usage decisions rest entirely with users and administrators. Different agent hosts apply varying constraints; do not assume hosts provide automatic protection. Third-party skill copyrights belong to their original authors. Use at your own risk.

1How It Works & Mechanism

Positioned as a "knowledge-based skill": AE is a GUI application where outputs are not rendered files, but pasteable assets (expressions, rigging schemes, export recipes, diagnostic workflows, naming conventions, and ExtendScript).

skills/after-effects/SKILL.md
After Effects is GUI software, so this skill delivers knowledge and ready-to-paste assets — expressions, rig setups, export checklists, diagnostic flows, naming schemes, and ExtendScript — rather than a rendered file.
What is being done here: Turning the reality that "an agent cannot click AE menus for you" into a design premise—deliverables are strictly limited to "text + runnable snippets," allowing the four references to carry far more depth than SKILL.md.

SKILL.md serves as a 4-domain routing hub: expressions/rigging/MOGRT, export recipes, performance triage, and project hygiene each route to a dedicated reference, keeping the root file minimal with actionable conclusions.

skills/after-effects/SKILL.md
| Write/debug an expression, rig sliders & nulls, build a MOGRT, export to Lottie | `references/expressions.md` (deep cookbook in `references/expression-library.md`) |
What is being done here: Replacing a bloated SKILL.md with a "what you need → which reference to consult" table to control resident context; this pattern is shared across all nine skills in this pack.

Expressions are ready to copy-paste immediately, grounded in AE expression evaluation semantics: JS evaluates on that property, value/time/index/thisLayer/thisComp are in scope, and expressions override keyframes by default.

skills/after-effects/SKILL.md
Expressions are JavaScript evaluated on the property they are pasted on; `value` is that property's value, `time` is comp time in seconds, and `index`, `thisLayer`, `thisComp` are in scope.
What is being done here: Establishing semantics before providing snippets—which is why it can confidently deliver wiggle(2, 30), loopOut("cycle"), and seedRandom(index, true) directly as answers.

Export confusion is traced back to "engine mismatch" (Render Queue = master/mezzanine, AME = delivery), establishing the black background issue caused by H.264 lacking an alpha channel as an architectural fact rather than a bug.

skills/after-effects/SKILL.md
Two engines, and choosing wrong is the root of most export confusion:
What is being done here: Breaking common support misdiagnoses into two distinct layers: "engine selection + codec capability"; two key conclusions (H.264 cannot store alpha; the H.264 output module was removed from RQ) are direct deductions from this.

Transparent exports require switching codecs (ProRes 4444 / PNG-TIFF sequences / HEVC-alpha + VP9-alpha) rather than tweaking matte settings.

skills/after-effects/SKILL.md
**H.264/MP4 cannot store alpha**, so a transparent comp exported to H.264 always fills with black; that is the classic "black background" complaint and not a bug.
What is being done here: Asserting "it's not a bug" stops users from futile parameter tweaking; the reference then provides step-by-step instructions and <video> tags for dual HEVC-alpha/WebM-alpha formats.

Performance triage is organized into three sequential symptom-based workflows (from least intrusive to most intrusive, stopping upon resolution) rather than a flat list of adjustable settings.

skills/after-effects/SKILL.md
Work top-down by symptom, cheapest fix first, and stop when it clears.
What is being done here: Splitting vague complaints like "AE is slow" first by symptom (slow preview, slow render, crash/freeze), organizing each into an ordered checklist with high-probability initial steps and exact preference paths.

Project hygiene establishes "delivery projects must Collect Files" and ".jsx batch operations must wrap in UndoGroup and apply effects by matchName" as hard rules.

skills/after-effects/SKILL.md
**Always hand off projects via `File > Dependencies > Collect Files`**, never a bare `.aep`.
What is being done here: These rules address two chronic pain points: "broken links across machines" and "localized display names breaking across language versions"; the matchName rule also serves as the implementation prerequisite for scripts/apply-effect-to-layers.jsx.

The three bundled .jsx scripts are templates intended for "editing top constants and running manually via Run Script in AE," not entry points for automated agent execution.

skills/after-effects/scripts/batch-rename.jsx
select layers in a comp, then File > Scripts > Run Script File...
What is being done here: Usage is documented in header comments (BASE_NAME / START_INDEX / PAD are configurable), and execution is always triggered manually by the user in AE—which is why risk assessment only evaluates local read/write.

2Core Capabilities

01Expressions and rigging: wiggle / loopOut / inertial bounce / seedRandom / valueAtTime echoes
02Controllable rigging and templates: null + Slider/Checkbox/Color controls, exported to .mogrt via Essential Graphics
03Export and encoding decisions: RQ vs AME, ProRes family, alpha channels, bitrates, AME preset pitfalls
04Local transcoding path for transparent video (ProRes 4444 master → ffmpeg VP9/WebM-alpha)
05Performance triage: 3 sequential workflows for slow preview / slow render / crashes and freezes + preference paths and recommended values
06Project hygiene: relinking (relinking one fixes a batch), Collect Files for portable delivery, naming conventions, and label color systems
07ExtendScript automation: batch rename / sequence staggering / batch effect application by matchName (3 bundled .jsx scripts)
08Lottie/Bodymovin export and boundaries (which features are reliable vs unsupported)

3External Dependencies

KindDependency / Endpoint
cliffmpeg(透明 WebM/VP9 转码示例)
packageAdobe After Effects 的两个导出引擎(Render Queue / Adobe Media Encoder)
packageBodymovin / Lottie 导出链(AE 扩展 + lottie-web;示例代码,产物期引入)
cliCreative Cloud 的 Other Versions 回滚流程(崩溃/回归时的用户侧动作)

4Recommended Governance & Evidence

Suggested Controls for Safe Operation
For Agent Runtime
Graded/per-use authorization, sandbox recommended
For Administrators
Approval workflow, rollback support, full audit
Evidence & Triggers (File Paths & Excerpts)
  • scripts/: Execute local scripts or commands: skills/after-effects/scripts/batch-rename.jsx (modifies the name of selected layers in the active comp)
  • 本地工作目录: File read/write and artifact output: .jsx scripts only manipulate in-memory AE project objects (layer names, in-points, effect stacks); disk persistence is determined by the user saving in AE; scripts do not write to the file system directly.
Audit Notes: Purely local behavior: documentation + expression snippets + 3 .jsx scripts operating exclusively on in-memory AE project objects; no external network calls, credential/environment variable access, file system I/O, or child processes. The sole execution surface is bundled scripts run manually by the user inside AE (matching the skill's declared deliverable format). Cautions (do not change rating tier): 1) References suggest enabling AE's "Allow Scripts to Write Files and Access Network" preference—a user-side host permission switch not used by the skill's own scripts; 2) Pasting untrusted expressions or .jsx files runs arbitrary ExtendScript inside AE, which is a source-trust issue rather than directory behavior.

5Second-Pass Verification

  • [ok] 三个 .jsx 是否含网络/文件系统/动态执行调用 — 逐行读完:只有 app.project.activeItem 类型守卫(CompItem/AVLayer)、sel[i].name、layers[j].startTime、layer.property("ADBE Effect Parade").canAddProperty/addProperty,以及 beginUndoGroup/endUndoGroup;无 XMLHttpRequest/fetch/File/socket/eval/Function。
  • [ok] SKILL.md 或 reference 里是否藏着会被 agent 执行的 shell/CLI 指令 — 全目录搜 npx/curl/wget/brew/npm:均无。本 skill 与同仓 color-motion 的关键差异正在这里——color-motion 的验收回路含 npx playwright screenshot,after-effects 没有任何需要跑的命令。
  • [ok] ffmpeg 示例是否构成「执行期外发」 — `ffmpeg -i master.mov … out.webm` 是纯本地转码,输入输出都在本机,不涉端点;列在 external_deps 只为交代工具链,不代表网络行为。
  • [ok] 「Allow Scripts to Write Files and Access Network」是否被本 skill 用来做事(是否该升档) — 该句出现在 references/project-hygiene.md 的 Scripting setup 段,是 AE 脚本能力的通用说明(AE 需开此开关脚本才能保存/渲染)。本 skill 的三个 .jsx 都不使用文件或网络 API;按判级口径(对象=skill 自身目录的程序行为),档位仍为蓝,该点只写入 risks。
  • [unlocatable] AE 行为类结论(QuickTime gamma、MFR 需每核数 GB、H.264 无 alpha)是否可在源码内验证 — 均为领域知识声明,pin 仓库内没有复现脚本、也没有引用来源(无外链、无数据集)。本次为只读侦查且不运行 AE,故不验证其数值正确性,只作为「文档声明」记录,不在结论里当作已核事实。

6Verdict & Scenarios

  • Deliverable formats align with software realities: AE is a GUI app, so the agent only provides knowledge and pasteable snippets—stated upfront without overpromising
  • Establishes two common misconceptions as definitive facts (H.264 cannot store alpha; RQ's H.264 output module is removed and requires AME), eliminating unproductive trial-and-error
  • Diagnostics follow sequential workflows rather than flat checklists: sorted by symptom, trying the least intrusive fixes first, stopping upon resolution, complete with exact preference paths and recommended values
  • Three .jsx scripts serve as minimal, directly runnable templates (UndoGroup + matchName + type guards), far more practical than pseudocode
  • Hard rules for cross-machine collaboration (Collect Files, relinking one fixes a batch) are highlighted prominently to avoid the classic ".aep only" disaster
  • Zero runtime external transmissions: the entire directory contains no network calls, credentials, or child processes; the security boundary is effectively limited to "the user running scripts manually"
  • Best for: Ideal for anyone working in Adobe After Effects: writing/debugging expressions, building controllable rigs and MOGRTs, deciding on export codecs (especially for transparency and black background issues), troubleshooting slow previews/renders/crashes, and organizing projects for clean delivery or collaboration. Also serves as an AE quick-reference guide—providing concise conclusions and menu paths first, with deep dives accessible in dedicated references.
    Not for: Not suited for: 1) Non-AE workflows (does not cover Remotion, GSAP, or other code-based animation engines; see remotion-video in this pack); 2) Expecting the agent to render output videos directly or automate the AE GUI (deliverables are strictly knowledge and snippets); 3) Complete solutions for advanced non-Lottie exports (such as 3D interchange formats)—only reliable Bodymovin boundaries are covered; 4) Cross-application pipeline integration (NLE timeline interchange, DaVinci Resolve color workflows), where this skill only provides codec and color tagging guidance.
    Install Agent-Ready
    ① Verified Mirror 2026-09-28
    Option A · Direct DownloadDownload after-effects.tar.gz
    sha256: 7cd8155a90d8ab34…
    Option B · Copy Guide for Agent
    install.json
    Agents automatically download and verify mirrors from this JSON guide.
    ② Upstream GitHub
    GitHub Repo ↗
    Pinned to commit 3c129f769d.