slack-gif-creator
Pass
Audited by Gen Agent Trust Hub on Sep 26, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted external data in the form of user-uploaded images to generate or inspire animations. This creates a potential surface for indirect prompt injection if an attacker embeds malicious instructions or metadata within an image file to influence the agent's behavior.
- Ingestion points:
SKILL.mdinstructs the agent to process user-uploaded files usingImage.open('file.png'). - Boundary markers: The instructions do not define specific delimiters or "ignore instructions" wrappers for the data derived from these images.
- Capability inventory: The skill possesses file-read (
Image.open) and file-write (builder.save) capabilities, along with image manipulation viaPIL. - Sanitization: There is no evidence of sanitization or metadata filtering for the ingested images.
Audit Metadata