agentcopv1.0.0

This plugin appears to do what it claims: it inspects all messages, tool calls, tool results, and outgoing HTTP requests and alerts on patterns that indicate injection, leaked credentials, SSRF/supply-chain calls, or excessive agency. Before installing, consider: (1) the plugin will see any sensitive data that passes through the agent (this is intentional for monitoring). If you handle secrets in conversations, be aware the detector inspects them for patterns; it does not send raw content out but it will log/deliver alerts containing detection metadata. (2) The only network call is to agentcop.live and only on the explicit /security badge command; if you trust that badge endpoint you can use it, otherwise avoid invoking the badge command. (3) The package contains tests that may try to call the badge endpoint if run — tests are not executed automatically at install but could perform network I/O in CI. (4) If you want quieter behavior, enable silentMode in the plugin config (logs only). Overall the code is coherent with its stated purpose; if you have strict data-flow or regulatory constraints, review how the agent handles sensitive inputs before enabling the plugin in production.