azsdk-common-pipeline-troubleshooting

Pass

Audited by Gen Agent Trust Hub on Jul 10, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill uses a set of specialized MCP tools (azure-sdk-mcp) to analyze and reproduce pipeline failures. The operations described (analyzing logs, verifying environments, and running local builds) are consistent with the stated goal of troubleshooting developer CI pipelines.
  • [INDIRECT_PROMPT_INJECTION]: The skill possesses an attack surface for indirect prompt injection as it processes external pipeline data.
  • Ingestion points: Pipeline logs and failure summaries are ingested via the azure-sdk-mcp:azsdk_analyze_pipeline tool as described in SKILL.md and evals/eval.yaml.
  • Boundary markers: There are no explicit instructions or delimiters defined to separate untrusted pipeline content from the agent's instructions.
  • Capability inventory: The skill has the ability to execute local build and validation commands through tools like azsdk_package_build_code and azsdk_package_run_check (SKILL.md).
  • Sanitization: No sanitization of ingested logs or validation of build scripts is mentioned in the provided documentation.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 10, 2026, 09:48 AM
Security Audit — agent-trust-hub — azsdk-common-pipeline-troubleshooting