hyperframes-video-beats
Pass
Audited by Gen Agent Trust Hub on Oct 18, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted transcript data to determine the timing, content, and placement of motion graphic beats. If a transcript contains malicious instructions or adversarial content, the agent may inadvertently follow them while generating or styling visual elements.
- Ingestion points: The skill explicitly instructs the agent to "Extract transcript chunks" and "Identify topic spans by narration" (SKILL.md).
- Boundary markers: There are no specific delimiters or instructions provided to the agent to disregard or escape potentially malicious instructions embedded within the transcript text.
- Capability inventory: The agent is authorized to use search tools (
rg,Select-String), execute Node/PowerShell scripts, run linter commands (npx hyperframes lint), and create or modify project files (compositions/beats/NNN-short-name.html,index.html). - Sanitization: No sanitization or validation logic is mentioned for the content extracted from transcripts before it is used to write UI text or logic.
- [COMMAND_EXECUTION]: The skill relies on the execution of several local command-line utilities and scripts to perform its tasks.
- Evidence: Instructions specify the use of
rg(ripgrep),Select-String, and "small Node/PowerShell transcript scripts" for data extraction. - Evidence: The skill requires running
npx hyperframes lintto validate the generated HTML and GSAP logic after editing files.
Audit Metadata