customize
Pass
Audited by Gen Agent Trust Hub on Sep 20, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill ingests user input to modify configuration files (
CLAUDE.mdandcompany-profile.md), creating a surface for indirect prompt injection where malicious instructions could be stored as configuration values. - Ingestion points: User input captured during the customization dialogue in Step 4.
- Boundary markers: None explicitly defined for the file-write operation.
- Capability inventory: File system write access to specific configuration paths (
~/.claude/plugins/config/...). - Sanitization: The skill includes logic to describe downstream impacts and check for internal inconsistencies before writing, providing a level of validation, though not explicit input sanitization.
Audit Metadata