SKILL.md
text/markdown · 3421 字节
rules/3d.md
text/markdown · 2290 字节
rules/animations.md
text/markdown · 790 字节
rules/assets/charts-bar-chart.tsx
text/plain · 3429 字节
rules/assets/text-animations-typewriter.tsx
text/plain · 2183 字节
rules/assets/text-animations-word-highlight.tsx
text/plain · 2387 字节
rules/assets.md
text/markdown · 1590 字节
rules/audio.md
text/markdown · 3822 字节
rules/calculate-metadata.md
text/markdown · 2920 字节
rules/can-decode.md
text/markdown · 1587 字节
rules/charts.md
text/markdown · 1751 字节
rules/compositions.md
text/markdown · 3445 字节
rules/display-captions.md
text/markdown · 3698 字节
rules/extract-frames.md
text/markdown · 5496 字节
rules/fonts.md
text/markdown · 3484 字节
rules/get-audio-duration.md
text/markdown · 1354 字节
rules/get-video-dimensions.md
text/markdown · 1627 字节
rules/get-video-duration.md
text/markdown · 1341 字节
rules/gifs.md
text/markdown · 3868 字节
rules/images.md
text/markdown · 2783 字节
rules/import-srt-captions.md
text/markdown · 2113 字节
rules/lottie.md
text/markdown · 1779 字节
rules/measuring-dom-nodes.md
text/markdown · 974 字节
rules/measuring-text.md
text/markdown · 3000 字节
rules/sequencing.md
text/markdown · 2389 字节
rules/tailwind.md
text/markdown · 422 字节
rules/text-animations.md
text/markdown · 700 字节
rules/timing.md
text/markdown · 3939 字节
rules/transcribe-captions.md
text/markdown · 850 字节
rules/transitions.md
text/markdown · 3730 字节
rules/trimming.md
text/markdown · 1210 字节
rules/videos.md
text/markdown · 3509 字节
skill-card.md
text/markdown · 2501 字节
Security Scan
状态
clean
OpenClaw
gpt-5.5
OpenClaw 分析
This is a Remotion documentation skill with relevant examples and no evidence of hidden, destructive, credential-stealing, or persistent behavior.
置信度: high
VirusTotal
Type: OpenClaw Skill Name: remotion-best-practices Version: 1.0.0 The skill bundle is primarily documentation and benign React/Remotion components. However, the `rules/tailwind.md` file contains a direct instruction to the AI agent: 'fetch https://www.remotion.dev/docs/tailwind using WebFetch for instructions.' While the target URL is legitimate documentation, this is an explicit prompt injection instructing the agent to perform a network call, which is a capability beyond the skill's stated purpose of providing best practices. This constitutes a risky capability without clear malicious intent, leading to a 'suspicious' classification.