audit-support

Pass

Audited by Gen Agent Trust Hub on Mar 23, 2026

Risk Level: SAFEPROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
  • [PROMPT_INJECTION]: The skill is vulnerable to Indirect Prompt Injection (Category 8) as it processes external, untrusted data from the '$ARGUMENTS' variable and scans local audit systems/evidence without implementing boundary markers or sanitization logic. This allows instructions embedded in the processed data to potentially influence the agent's behavior. (Evidence Chain: Ingestion points: $ARGUMENTS and audit documents/GRC systems; Boundary markers: Absent; Capability inventory: File creation, directory creation, and appending to telemetry logs; Sanitization: Absent).
  • [COMMAND_EXECUTION]: The skill performs automated file system operations, including creating a 'docs/' directory and writing report files. It also implements a 'Self-Evolution Telemetry' mechanism that reads and appends to files within the hidden '~/.claude/projects/' directory to track execution metadata.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 23, 2026, 10:56 AM
Security Audit — agent-trust-hub — audit-support