audit-payment-system

Pass

Audited by Gen Agent Trust Hub on Aug 25, 2026

Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill instructs the agent to run rg (ripgrep) shell commands in SKILL.md (Phase 0) and references/checklist.md to scan the repository for payment providers, money-handling logic, and sensitive data patterns. These are read-only search operations and do not represent arbitrary command injection.
  • [PROMPT_INJECTION]: The skill analyzes untrusted third-party source code and logs, which presents a surface for indirect prompt injection. Malicious instructions embedded in the audited files could potentially influence the agent's reporting or behavior.
  • Ingestion points: Source code files, database schemas, and application logs scanned via rg commands in Phase 0 and Phase 2.
  • Boundary markers: The instructions do not define delimiters or provide specific warnings to the agent to ignore instructions found within the analyzed data.
  • Capability inventory: Uses rg for searching; generates text-based audit reports. No file-write or external network capabilities are initiated by the skill scripts themselves.
  • Sanitization: No evidence of escaping or filtering of content retrieved from the audited files before it is processed by the agent.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 25, 2026, 05:55 AM
Security Audit — agent-trust-hub — audit-payment-system