azure-ai-contentunderstanding-py

Pass

Audited by Gen Agent Trust Hub on Feb 14, 2026

Risk Level: LOWEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
  • Indirect Prompt Injectio n (LOW): The skill ingests untruste d external media which may contain maliciou s instruction s designe d to influenc e the agent.
  • Ingestio n points: The client.begin_analyze metho d in SKILL.md processe s conten t from untruste d URLs via AnalyzeInput.
  • Boundary markers: Absent; the skill extract s raw markdow n and transcrip t s withou t delimiter s or 'ignor e' instruction s.
  • Capabilit y inventor y: The skill extract s text ual conten t but lack s privilege d operation s like file write s or subproces s executio n.
  • Sanitizatio n: Absent; raw extracte d conten t is passe d directl y to the cons ume r.
  • External Dow nloa d s (LOW): The skill install s azure-ai-content understanding, a verifie d packag e from Microsof t, which is a truste d source per [TRUST-SCOPE-RULE].
Audit Metadata
Risk Level
LOW
Analyzed
Feb 14, 2026, 04:58 PM