privacy-review-rails
Pass
Audited by Gen Agent Trust Hub on Mar 26, 2026
Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill executes a local Ruby script (
scanner.rb) from a sibling directory (privacy-by-design-rails) to perform automated privacy scanning. It also uses dynamic context injection (!test -d ...) to verify the existence of this dependency at load time. - [PROMPT_INJECTION]: The skill has a surface for indirect prompt injection because it reads and processes untrusted project files (Ruby code, YAML, migrations) which could contain embedded instructions designed to manipulate the agent's behavior.
- Ingestion points: Project files identified by the scanner as
changed_filesorrelated_files(SKILL.md). - Boundary markers: None present to delimit untrusted file content from instructions.
- Capability inventory: The skill utilizes
Bash,Edit,Read,Glob, andGreptools, which allow for file system modification and command execution. - Sanitization: No sanitization or escaping of the ingested file content is performed before processing.
Audit Metadata