report-generation
Pass
Audited by Gen Agent Trust Hub on Sep 17, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes data from external sources (JSON, CSV, YAML, API responses) which may be untrusted.
- Ingestion points: Workflow Step 1 defines the entry of external data payloads into the agent context.
- Boundary markers: The instructions do not define delimiters or specific warnings for the ingested data.
- Capability inventory: The skill can generate Markdown, HTML, and PDF files, and produces chart specifications.
- Sanitization: The instructions mention schema validation and PII redaction, but do not specify sanitization of data fields to prevent malicious prompt instructions.
Audit Metadata