review
Pass
Audited by Gen Agent Trust Hub on Aug 27, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTIONDATA_EXFILTRATION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted project source code and documentation, which could contain instructions designed to manipulate the audit process. Ingestion points include all audited source files and documentation. Boundary markers are absent, providing no separation between data and instructions. The skill has capabilities to read/write files and execute shell commands. No sanitization is performed on input files before processing.
- [COMMAND_EXECUTION]: The skill executes shell commands for static analysis based on definitions in the local
docs/standards/tech-stack-rule.mdfile. Evidence inSKILL.mdandtemplates/review-global-prompt.mdshows the agent running these user-provided commands, which could lead to arbitrary execution if the config file is compromised. - [DATA_EXFILTRATION]: The skill identifies and documents hardcoded credentials as part of its security audit.
templates/review-global-prompt.mdincludes 'hardcoded sensitive information' in its checklist. While intended for security improvement, discovered secrets are written to local reports which could result in data exposure.
Audit Metadata