AI & machine learning Plugins

Choose by task and evidence

A plugin is a package for a specific agent client. It can bundle instructions, commands, hooks, application integrations or MCP connections, so begin by identifying the capability you actually need. Compare the detected manifest format with your client and inspect the package components before installing it. Start with a source-based review, then use a disposable project to check loading, permissions and the expected result. Repository activity and a declared license help explain the evidence; they do not certify the behavior of bundled code or connected services. Avoid adding a second connection when the plugin already provides it.

Agent labels reflect detected manifests or repository setup documentation. The runtime of each entry has not been tested in every client. Use the evidence links to check your exact client and version.