ecosystem-health

Pass

Audited by Gen Agent Trust Hub on Apr 30, 2026

Risk Level: SAFE
Full Analysis
  • [COMMAND_EXECUTION]: The skill uses standard CLI tools such as find, ls, grep, cat, jq, and python3 to perform diagnostic scans of the local filesystem and configuration files (e.g., ~/.claude.json). These operations are consistent with the skill's stated purpose of ecosystem health monitoring.
  • [EXTERNAL_DOWNLOADS]: The skill invokes npm outdated to check for updates for local MCP servers. This is a routine package management check and is limited to metadata retrieval rather than execution of untrusted remote code.
  • [DATA_EXFILTRATION]: While the skill accesses potentially sensitive configuration files and user-defined vault paths, the gathered information is used only to generate a local health report. There are no network operations or exfiltration attempts identified in the implementation.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 30, 2026, 12:59 PM
Security Audit — agent-trust-hub — ecosystem-health