sparkscan-net-ios
Pass
Audited by Gen Agent Trust Hub on Aug 1, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill fetches the latest SDK version information from NuGet, a well-known package registry.
- [PROMPT_INJECTION]: The skill contains instructions that modify the agent's default transparency, such as directing the agent to modify files directly instead of displaying a preview in chat ("Do not just show the code in chat; apply it to the file") and suppressing certain suggestions ("Do not tell the user to check the docs themselves"). These instructions are intended for user experience but technically override standard agent behavior.
- [PROMPT_INJECTION]: The skill presents an attack surface for indirect prompt injection due to its design to ingest and analyze existing user code from project files while maintaining the capability to perform file-write operations. Ingestion points: Instructions in
references/migration.mdandreferences/third-party-migration.mdrequire the agent to analyze user-provided code. Boundary markers: There are no explicit markers or instructions to ignore instructions embedded within the ingested code. Capability inventory: The agent is authorized to modify user files. Sanitization: There is no defined process for validating or sanitizing external code before modification.
Audit Metadata