optimize-agent-capabilities
Pass
Audited by Gen Agent Trust Hub on Jul 23, 2026
Risk Level: SAFECOMMAND_EXECUTION
Full Analysis
- [COMMAND_EXECUTION]: The script
scripts/audit.mjsusesnode:child_process.spawnSyncto executesh -lc command -vcommands. This is used solely to verify the existence of agent CLI tools (e.g., 'claude', 'kimi', 'codex') within the user's shell environment to determine which clients are active. - [SAFE]: The skill performs a local filesystem audit of specific AI agent configuration paths (e.g.,
~/.claude,~/.codex,~/.kimi) and counts the number of skill files found. These operations are consistent with the skill's stated purpose of auditing agent capabilities and do not access sensitive credentials, private keys, or perform unauthorized data exfiltration. - [SAFE]: The skill instructions and scripts consistently emphasize a read-only approach, explicitly forbidding the installation, modification, or deletion of files and client configurations.
Audit Metadata