Skills motion-graphics

📦

motion-graphics

Name: motion-graphics
Author: heygen-com

Content revision r2 High Risk 📁 Filesystem access🔑 Env variables⚙️ External commands🌐 Network access

Create polished short motion graphics

Short motion graphics can be difficult to plan, build, and verify consistently. This skill guides a structured workflow for design-led videos, overlays, charts, maps, and branded animations.

Supports: Claude Codex Code(CC)

⚠️ 38 Poor

Stop for confirmation before installing.

Review the plan and obtain explicit user consent before changing files.

Install with my Agent

Copy this request to your Agent. It includes the canonical Skill page and manifest.

Agent request

Review the Skillstore skill "motion-graphics" from https://skillstore.io/skills/heygen-com-motion-graphics.md and its manifest at https://skillstore.io/api/skills/heygen-com-motion-graphics/manifest. Verify the artifact. Stop and obtain explicit user consent before installing or changing files.

Your Agent should still show its plan and request any confirmation required by the security policy.

Agent-readable resources

Use these links when an AI agent, crawler, or script needs clean context instead of reading the full page.

Markdown detail GET /skills/heygen-com-motion-graphics.md Signed manifest GET /api/skills/heygen-com-motion-graphics/manifest Signed lockfile GET /api/skills/heygen-com-motion-graphics/lockfile

Test it

Using "motion-graphics". Create a five-second animated statistic for 75 percent adoption.

Expected outcome:

A short count-up animation from 0 to 75 percent.
A final hold with the metric and a concise label.
A rendered video file and proof snapshots.

Using "motion-graphics". Create a branded lower third for Maya Chen, Product Director.

Expected outcome:

A reusable lower-third composition using the supplied brand colors.
An entrance, readable hold, and exit animation.
A rendered overlay-ready video when the requested format is available.

Security Audit

High Risk

v6 • 7/19/2026 Open versioned report

The skill contains legitimate motion-graphics tooling, but its map and localization helpers execute local programs, access remote resources, and handle an optional API credential. Review found unsafe autonomous update guidance, an unsandboxed browser with remote scripts, and missing output-path and external-content trust controls. Most template-literal, documentation, and visual-animation matches were false positives.

Files scanned

1,660

Lines analyzed

Review items

False positives ignored

Confirmed security concerns (5)

High

grounding/locate.mjs:191

Environment file access

const key = process.env.GEMINI_API_KEY;

The code accesses GEMINI_API_KEY from the environment. It is an actual credential-handling path, despite the pattern name.

High

categories/maps/bake-basemap.mjs:117-119 categories/maps/bake-basemap.mjs:184-196

Unsandboxed browser executes remote content

The map baker disables the Chrome sandbox and loads JavaScript and map resources from remote endpoints. A compromised remote dependency or browser exploit could affect the host build environment.

The source explicitly loads remote scripts and passes --no-sandbox to Puppeteer. These controls combine into a clear host-isolation risk.

High

SKILL.md:22-24 SKILL.md:107-122

External content prompt injection exposure

The autonomous workflow sources external assets and includes resolved asset information in subagent context without an explicit rule that fetched content is untrusted data. Malicious webpages or assets can contain instruction-like content.

The workflow is autonomous and forwards resolved external assets to subagents. The reviewed instructions do not establish a content trust boundary.

Medium

categories/maps/bake-basemap.mjs:61 categories/maps/bake-basemap.mjs:111-113 categories/maps/bake-basemap.mjs:227-257

Output path traversal through environment name

The map baker uses the unvalidated NAME environment value to form frame directories and output filenames. A value containing path separators can escape the intended output directory.

NAME is directly concatenated into paths without basename validation or containment checks.

Medium

SKILL.md:14 SKILL.md:90

Unattended remote package update

The skill directs the agent to run a silent npx update before work begins, allowing remotely fetched package code to change the installed skill without confirmation.

The instructions explicitly say to update silently and describe global updates from GitHub.

Capability review items (23)

These are real local capabilities that may be expected for this skill, so they require review but are not counted as confirmed malicious behavior.

High

categories/maps/bake-basemap.mjs:25

Node.js child_process module

import { spawnSync } from "node:child_process";

The helper launches a local browser process to render map frames. It is a real execution capability and the launched browser later disables its sandbox.

High

categories/maps/bake-basemap.mjs:229

Synchronous spawn

const ff = spawnSync(

The helper runs ffmpeg with filesystem-derived output paths and overwrite enabled. This is a real process-execution capability.

High

grounding/locate.mjs:23

Node.js child_process module

import { execFileSync } from "node:child_process";

The localizer invokes external media tools. Although it uses argument arrays, it is a real local command-execution capability.

High

grounding/locate.mjs:35

Synchronous file execution

execFileSync("ffmpeg", ["-v", "error", "-y", ...args], {

The localizer executes ffmpeg to transform caller-supplied image files. This is intentional but requires trusted inputs and a controlled environment.

High

grounding/locate.mjs:39

Synchronous file execution

const out = execFileSync("ffprobe", [

The localizer executes ffprobe on caller-supplied image files. This is intentional but is a real external execution capability.

High

grounding/locate.mjs:199

Synchronous file execution

const b64 = execFileSync("base64", ["-i", img]).toString().replace(/\n/g, "");

The optional detector path executes base64 on the selected image before upload. This is a real command execution capability.

High

grounding/locate.mjs:191

Generic API/secret keys

const key = process.env.GEMINI_API_KEY;

The optional detector reads GEMINI_API_KEY and transmits it to a remote API endpoint.

High

grounding/locate.mjs:194

Generic API/secret keys

"auto needs GEMINI_API_KEY; use the grid loop instead (overlay→region→final→mark)",

The documentation and error message expose that the optional auto mode requires GEMINI_API_KEY.

High

grounding/PROTOCOL.md:19

Generic API/secret keys

1. **A strong detector is available** (e.g. `GEMINI_API_KEY` in env) →

The protocol directs use of an environment API key for the optional detector path.

Medium

categories/maps/bake-basemap.mjs:257

Synchronous file operations

writeFileSync(join(OUT, NAME + "-coords.json"), JSON.stringify(coords));

The helper writes a coordinate JSON artifact using an environment-controlled output name. Unvalidated path components can affect the write destination.

Medium

SKILL.md:14

Ruby/shell backtick execution

> **First, keep this skill fresh — run silently, don't ask:** `npx hyperframes skills update motion-

The workflow instructs an unattended npx skill update before use. This can execute newly fetched package code without user confirmation.

Low

categories/maps/bake-basemap.mjs:153

Fetch API call

fetch("https://cdn.jsdelivr.net/npm/world-atlas@2.0.2/countries-110m.json").then(function(r){return

The generated browser page fetches remote geographic data at runtime. This creates a network and third-party content trust boundary.

Low

categories/maps/bake-basemap.mjs:106

Hardcoded URL

"https://server.arcgisonline.com/ArcGIS/rest/services/World_Imagery/MapServer/tile/{z}/{y}/{x}",