The Agent Skills Directory

[PROMPT_INJECTION]: The skill instructions in SKILL.md focus on guiding the agent to select the correct animation category based on user descriptions. It does not contain any instructions attempting to bypass safety filters or override system constraints.
[DATA_EXFILTRATION]: No sensitive file paths, credential harvesting patterns, or suspicious network operations were detected. The skill primarily reads its own internal reference files (ref/*.md) and outputs text to the user.
[REMOTE_CODE_EXECUTION]: The skill provides UI code snippets for the user to implement but does not execute arbitrary remote code itself. The installation instructions provided in README.md use standard, well-known methods (npx, git clone).
[OBFUSCATION]: No malicious obfuscation techniques such as multi-layer Base64, zero-width characters, or homoglyphs were found. Standard Data URIs are used for the favicon in the documentation files, which is normal practice for web development.
[COMMAND_EXECUTION]: The skill does not use dynamic shell execution or dangerous subprocess patterns. There are no instances of the dynamic context injection syntax (!command) in the SKILL.md file.
[INDIRECT_PROMPT_INJECTION]: While the skill processes user input to recommend animations, it lacks the execution capabilities (like writing to files or calling shell commands) that would make it vulnerable to indirect injection attacks. The risk surface is minimal.

motion-ref