agent-capability-analyzer
Pass
Audited by Gen Agent Trust Hub on Oct 3, 2026
Risk Level: SAFECOMMAND_EXECUTIONDATA_EXFILTRATION
Full Analysis
- [COMMAND_EXECUTION]: The skill uses Node.js scripts to automate data collection and management. This includes
populate-agent-descriptions.mjs, which usesexecFileSyncto call the update script for each agent, andupdate-agent-map.mjs, which manages a local SQLite database using the built-innode:sqlitemodule. These operations are limited to project-local metadata management. - [DATA_EXFILTRATION]: While the skill involves collecting data from multiple agents and 'dumping' it to a JSON file, the destination is strictly local (
.plugin-creator/audits/agent-map.json). There are no network operations, and data remains within the project's security boundary. - [SAFE]: The skill's primary purpose is auditing and governance. It provides a structured way to identify 'drift' or 'scope violations' in agent behavior, which is a positive security and reliability practice. The use of standard tools like
pnpmandlevelis consistent with development workflows.
Audit Metadata