kanchi-dividend-us-tax-accounting
Pass
Audited by Gen Agent Trust Hub on Oct 6, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill ingests user-controlled financial data which serves as a potential surface for indirect prompt injection attacks.
- Ingestion points: Data such as ticker symbols, instrument types, and holding periods are ingested via a JSON input file processed by the
scripts/build_tax_planning_sheet.pyutility or through natural language interaction. - Boundary markers: The skill documentation defines a structured JSON schema but does not provide the agent with explicit instructions or delimiters to disregard potential instructions embedded within these data fields.
- Capability inventory: The skill possesses the capability to read local files and write Markdown and CSV reports to a specified output directory using the provided Python script. It does not request network access or privilege escalation.
- Sanitization: The
build_tax_planning_sheet.pyscript performs basic data normalization, including type-casting and case-normalization, which provides limited sanitization against malformed data but does not inspect for malicious instructions.
Audit Metadata