framer-motion
Warn
Audited by Gen Agent Trust Hub on Sep 22, 2026
Risk Level: MEDIUMMETADATA_POISONINGINDIRECT_PROMPT_INJECTION
Full Analysis
- [METADATA_POISONING]: The file skill-report.json attributes authorship to 'Asmayaseen', which conflicts with the authoritative vendor context of 'aiskillstore', representing a deceptive metadata pattern.
- [INDIRECT_PROMPT_INJECTION]: The skill includes a skill-report.json file containing self-referential safety verdicts (e.g., 'risk_level': 'safe', 'safe_to_publish': true). This attempts to override the analyzer's judgment and influence the final security verdict.
- [INDIRECT_PROMPT_INJECTION]: The skill defines several React templates that ingest external content. 1. Ingestion points: 'children', 'items', and 'renderItem' props in templates/animated-list.tsx, templates/page-transition.tsx, and examples/list-animations.md. 2. Boundary markers: Absent. 3. Capability inventory: Analysis of all skill files indicates zero subprocess calls, network operations, or dangerous file system writes. 4. Sanitization: The templates lack explicit sanitization or validation of the ingested React node data.
Audit Metadata