aso
Pass
Audited by Gen Agent Trust Hub on Nov 5, 2026
Risk Level: SAFEMETADATA_POISONINGINDIRECT_PROMPT_INJECTION
Full Analysis
- [METADATA_POISONING]: The skill's instructions contain numerous specific references to fictional future events, versions, and products (e.g., WWDC25, iOS 26/27, iPhone Duo, and App Review Guideline updates from late 2026). While these provide a detailed persona for the AI, they represent misleading information presented as factual guidance, though they do not impact system safety.\n- [INDIRECT_PROMPT_INJECTION]: The skill describes processes for auditing and reviewing user-provided metadata (app names, descriptions, keywords), which introduces an attack surface where untrusted external data is ingested into the agent context.\n
- Ingestion points: External App Store metadata provided by the user for auditing or optimization (e.g., app descriptions, competitor metadata).\n
- Boundary markers: None identified; the skill does not instruct the agent to use specific delimiters or to ignore embedded instructions in the metadata being audited.\n
- Capability inventory: The skill is purely instructional and does not request access to external tools, shell execution, or network operations, limiting the impact of any potential injection.\n
- Sanitization: No sanitization or validation of the input metadata is specified.
Audit Metadata