create-report
Pass
Audited by Gen Agent Trust Hub on Sep 11, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted data from local files and web research to populate HTML reports, creating a surface for potential instruction injection through external content.
- Ingestion points: SKILL.md (Workflow step 2) describes gathering and verifying sources from supplied files and output from the web-research tool.
- Boundary markers: The skill requires strict labeling of evidence (Observed, Source-reported, Inference, Unknown) as defined in references/report-design.md to distinguish between different levels of claim certainty.
- Capability inventory: The skill utilizes local file-write access to create the HTML artifact and browser automation for QA validation as per compatibility requirements in SKILL.md.
- Sanitization: The skill mandates HTML escaping for all inserted content, redaction of secrets and PII, and employs a highly restrictive Content Security Policy (CSP) in the assets/report-template.html to prevent unauthorized network requests or data exfiltration.
Audit Metadata