dinglebear-ai/labby
Labby gateway skills and MCP client integration.
Implement a repository task in a persistent Microsandbox, preserve research and test evidence, publish an exact Git commit through an authorized path, and retain a separate verified staging instance without touching production.
Use when installing, configuring, repairing, or validating Labby, including auth, deployment, public exposure, and agent MCP setup.
Use when discovering and calling upstream MCP tools, prompts, resources, or skills through Labby Code Mode; batching gateway calls; receiving and acknowledging response notifications; or diagnosing Code Mode scope, schema, and execution errors. For saved reusable workflows, use using-snippets.
Use when operating an installed Labby through its CLI, MCP, HTTP API, or web UI; updating it, checking health and logs, managing gateway upstreams and configuration, or exporting and restoring durable state. For live Code Mode calls use using-codemode; for saved snippets use using-snippets.
Use when discovering, running, creating, editing, promoting, validating, testing, explaining, or removing Labby Code Mode snippets; or when turning a successful Code Mode execution into a reusable workflow.