opusclip
Pass
Audited by Gen Agent Trust Hub on Mar 1, 2026
Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill's CLI script invokes system commands like
openorxdg-opento display generated HTML files in a web browser. This occurs when thepreviewcommand is executed. - [PROMPT_INJECTION]: The skill exhibits an indirect prompt injection surface during the generation of video previews.
- Ingestion points: Data is retrieved from the OpusClip API (e.g., clip titles, descriptions) and used in the
cmd_previewfunction within thescripts/opusclipscript. - Boundary markers: No delimiters or safety instructions are present to isolate external API content from the HTML structural elements.
- Capability inventory: The skill utilizes
curlfor network access, writes local files to the/tmpdirectory, and launches system browsers through shell commands. - Sanitization: The script lacks escaping or sanitization when inserting API-derived strings into the HTML preview template. It relies on simple bash string substitution (
html="${html//\{\{CLIP_CARDS\}\}/$cards}") which could allow for the injection of malicious HTML elements or scripts from the API response into the local user environment.
Audit Metadata