diag-harness

Pass

Audited by Gen Agent Trust Hub on Aug 14, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill implements a diagnostic reporting feature (--bundle) that includes a built-in safety mechanism to redact sensitive information. It explicitly filters out object keys matching patterns for secrets, tokens, passwords, and API keys before outputting data intended for public support tickets.
  • [COMMAND_EXECUTION]: The skill detects version skews and provides transparent recommendations for standard package management commands (npm install) to align dependencies. These commands are descriptive and intended for manual user execution.
  • [EXTERNAL_DOWNLOADS]: The skill references standard ecosystem packages under the @metaharness namespace. These references are part of the core functionality for version validation and dependency resolution within the tool's intended framework.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 14, 2026, 07:38 PM
Security Audit — agent-trust-hub — diag-harness