cursor-multi-repo
Pass
Audited by Gen Agent Trust Hub on Sep 14, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill facilitates the ingestion of data from multiple repository roots, which could contain malicious configuration or instruction files.\n
- Ingestion points: The skill reads repository structures and configuration files, such as
.cursorignore,.cursor/rules/, and.code-workspacefiles across multiple repository roots.\n - Boundary markers: Absent. There are no explicit instructions to distinguish between the skill's instructions and potentially malicious instructions embedded within the project files.\n
- Capability inventory: The skill is configured with
Read,Write, andEditpermissions, along with arbitrary command execution viaBash(cmd:*).\n - Sanitization: Absent. No sanitization or validation of the content within the indexed files or rules is described.
Audit Metadata