adobe-after-effects
Pass
Audited by Gen Agent Trust Hub on Oct 23, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONDYNAMIC_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and analyze untrusted data, specifically expression text and screenshots of the After Effects interface, to perform diagnostic tasks.
- Ingestion points:
gemini.mddescribes a workflow for analyzing user-provided screenshots of error banners and status panels. - Boundary markers: The skill includes operational directives for the agent to request specific evidence (e.g., "request expression text + layer/property path screenshot") before providing a fix.
- Capability inventory: The skill can generate and write files (ExtendScript and Python) and manipulate project properties through the After Effects DOM.
- Sanitization: Directives emphasize verifying settings and paths (e.g., "confirm Output Module path includes container extension") rather than blindly trusting inputs.
- [DYNAMIC_EXECUTION]: The skill includes a Python script in
gpt.mdthat generates ExtendScript (.jsx) files for local use. - This functionality is used for boilerplate code generation, such as batch renaming compositions.
- The script uses static templates and does not dynamically execute code or fetch untrusted remote content.
Audit Metadata