CPA
Pass
Audited by Gen Agent Trust Hub on Aug 13, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill's instructions and metadata do not contain any malicious patterns, obfuscation, or unauthorized access attempts. All functionalities described are consistent with its stated purpose as a study assistant.
- [DATA_EXPOSURE_AND_EXFILTRATION]: The skill manages candidate data locally in the
~/Clawic/data/cpa/directory. No external network operations, hardcoded credentials, or exfiltration patterns were detected. The use of a dedicated local directory for state persistence is consistent with safe development practices. - [INDIRECT_PROMPT_INJECTION]: The skill is designed to parse external documentation, such as CPA score reports and educational transcripts, which introduces a potential surface for indirect prompt injection.
- Ingestion points: External score reports, credential evaluations, and user-provided education history files.
- Boundary markers: No specific delimiters or "ignore embedded instructions" warnings are defined in the current prompt instructions.
- Capability inventory: The skill tracks progress and provides advice but does not include scripts for arbitrary code execution, subprocess calls, or network requests.
- Sanitization: The instructions do not specify validation or sanitization routines for the content of score reports.
Audit Metadata