validate-harness

Pass

Audited by Gen Agent Trust Hub on Aug 14, 2026

Risk Level: SAFECOMMAND_EXECUTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill facilitates the execution of the harness CLI tool, the gcloud CLI, and local Node.js scripts. These commands are used to verify authentication states and environment configurations within the local development context without evidence of external exfiltration.
  • [INDIRECT_PROMPT_INJECTION]: The skill ingests and analyzes content from local source files (TypeScript, JavaScript, Rust) and JSON configuration files. This represents an ingestion surface where local project data is processed as part of the validation logic.
  • Ingestion points: Local project source code and .mcp/servers.json files.
  • Boundary markers: No specific delimiters or safety instructions are provided to the agent to distinguish between file content and instructions.
  • Capability inventory: File system read access, execution of CLI validation tools, and GCP Secret Manager metadata access.
  • Sanitization: The instructions do not describe any sanitization, escaping, or filtering of the file contents prior to analysis.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 14, 2026, 07:38 PM
Security Audit — agent-trust-hub — validate-harness