powershell-expert

Pass

Audited by Gen Agent Trust Hub on Aug 15, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • [PROMPT_INJECTION]: The skill includes a skill-report.json file containing self-referential safety claims (e.g., "risk_level: safe", "safe_to_publish: true") that target the evaluation process rather than providing technical evidence.
  • [PROMPT_INJECTION]: The skill has a surface for indirect prompt injection due to its processing of external content.
  • Ingestion points: The agent is instructed to fetch data from the PowerShell Gallery and Microsoft Docs via WebFetch in SKILL.md (Step 1 and Step 2).
  • Boundary markers: The fetching instructions lack explicit delimiters or warnings to ignore instructions embedded in the external content.
  • Capability inventory: The agent can execute a bundled PowerShell script (Search-Gallery.ps1) and generate code for the user based on external inputs.
  • Sanitization: The skill uses basic field extraction but does not sanitize the full prose content of external package descriptions or documentation.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 15, 2026, 01:57 AM
Security Audit — agent-trust-hub — powershell-expert