motion-graphics
Pass
Audited by Gen Agent Trust Hub on Aug 21, 2026
Risk Level: SAFEREMOTE_CODE_EXECUTIONCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [REMOTE_CODE_EXECUTION]: The skill uses
npxto manage its operating environment, including updating its own skills (npx hyperframes skills update) and initializing projects. These operations execute packages fetched from the public npm registry at runtime. - [COMMAND_EXECUTION]: The skill invokes system-level binaries to handle media operations. Scripts such as
categories/maps/bake-basemap.mjsandgrounding/locate.mjsutilizespawnSyncandexecFileSyncto driveffmpegandffprobefor video encoding and image analysis. - [INDIRECT_PROMPT_INJECTION]: The skill ingests data from external sources which constitutes a potential attack surface.
- Ingestion points: The skill fetches content from news articles, tweets, and web pages during the sourcing phase as defined in
phases/source/guide.md. - Boundary markers: The skill lacks explicit boundary markers or instructions to differentiate between trusted agent instructions and untrusted data during the planning phase in
agents/director.md. - Capability inventory: The skill can write arbitrary HTML/JS files to the local file system and trigger video rendering via shell commands.
- Sanitization: No visible sanitization or validation is applied to the ingested search results before they are used to generate keywords, layout, and content for the motion graphic compositions.
- [DYNAMIC_CONTEXT_INJECTION]: The
SKILL.mdfile contains a directive using backtick syntax (npx hyperframes skills update motion-graphics) designed to be executed silently by the platform when the skill is loaded.
Audit Metadata