Skills
skills/bobmatnyc/claude-mpm-skills/gh-cli/Gen Agent Trust Hub

gh-cli

Warn

Audited by Gen Agent Trust Hub on Jun 16, 2026

Risk Level: MEDIUMCOMMAND_EXECUTIONSAFE
Full Analysis
  • [COMMAND_EXECUTION]: The skill provides patterns for using aws ssm send-command to execute shell commands on remote AWS instances. This is a powerful administrative capability intended for robust automation and CI environments.
  • [COMMAND_EXECUTION]: To resolve authentication issues where the GitHub CLI cannot access the OS keychain, the skill instructs the agent to disable process sandboxing (e.g., using the dangerouslyDisableSandbox: true configuration). This grants the tool direct access to sensitive credential stores on the host system.
  • [COMMAND_EXECUTION]: Instructions include manipulating the GH_CONFIG_DIR environment variable and searching the file system for hosts.yml to identify and point to valid GitHub credential directories.
  • [SAFE]: The skill implements standard GitHub operations (issue, PR, and label management) using the official gh CLI tool.
  • [SAFE]: All external references and service patterns target well-known, trusted platforms including GitHub and Amazon Web Services (AWS).
Audit Metadata
Risk Level
MEDIUM
Analyzed
Jun 16, 2026, 04:19 PM
Security Audit — agent-trust-hub — gh-cli