azsdk-common-prepare-release-plan

Pass

Audited by Gen Agent Trust Hub on Jul 10, 2026

Risk Level: SAFE
Full Analysis
  • [DATA_EXPOSURE]: The skill includes explicit instructions in SKILL.md and references/release-plan-details.md to avoid displaying internal Azure DevOps work item URLs to the user, providing only the public Release Plan Link and ID instead. This is a proactive measure against accidental internal data exposure.
  • [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted user input including TypeSpec project paths and GitHub PR URLs which are then passed to MCP tools. While there are no explicit boundary markers or sanitization steps mentioned, the capabilities are limited to specific release management tools within a trusted environment (Azure SDK infrastructure).
  • Ingestion points: User-provided TypeSpec project paths and GitHub PR URLs are used as parameters in SKILL.md and references/release-plan-details.md.
  • Boundary markers: None identified.
  • Capability inventory: The skill invokes several azure-sdk-mcp tools to create, update, and link release data.
  • Sanitization: No explicit sanitization or validation of the input strings is performed beyond verifying the repository names.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 10, 2026, 09:48 AM
Security Audit — agent-trust-hub — azsdk-common-prepare-release-plan