cursor-setup-pstack
Pass
Audited by Gen Agent Trust Hub on Oct 19, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill references and reads configuration from the vendor's repository at 'TheAngryPit/TheAngrySkills' and the official Cursor plugins repository. These downloads are directed at legitimate vendor infrastructure or well-known services and are used to synchronize model routing policies.
- [INDIRECT_PROMPT_INJECTION]: The skill ingests data from external routing policies and remote GitHub sources to determine model capabilities and role mappings. Although this represents an ingestion surface for potentially untrusted data, the skill mitigates risks by requiring reviewed diffs for updates and providing explicit validation steps to ensure proposed models are supported by the execution channel.
- [DYNAMIC_EXECUTION]: The skill can generate project-specific verification tools, such as formal verification skills, upon user request. This code generation is handled as an optional, interactive step that requires explicit operator consent before any artifacts are created.
Audit Metadata