remove-animations
Pass
Audited by Gen Agent Trust Hub on Oct 28, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted repository content (source code, CSS, and configuration files) which acts as an ingestion point for external data.
- Ingestion points: The skill reads various project files, including
.tsx,.css, and Tailwind configurations, to identify animation patterns. - Boundary markers: The skill contains an explicit security rule (Hard Rule 5) instructing the agent to treat repository content strictly as data and to ignore/flag common injection patterns like "ignore previous instructions".
- Capability inventory: The agent has the capability to edit and delete local code files.
- Sanitization: The skill implements defensive prompting to ensure instructions embedded in data are not executed, which is a recommended practice for skills handling untrusted codebases.
- [PROMPT_INJECTION]: A static analysis hint flagged a potential instruction override pattern.
- Evidence: The string "ignore previous instructions" appears in the text.
- Context: This string is used as an example in a defensive rule meant to protect the agent. It instructs the agent to ignore such phrases if they are encountered in the files being analyzed. This is a benign, safety-oriented use of the pattern.
Audit Metadata