GitHub Copilot Launches Agent Finder for Dynamic Resource Discovery
Key Takeaways
- ▸Agent Finder enables GitHub Copilot agents to dynamically discover and load resources on-demand, eliminating the need for manual pre-configuration and context window bloat
- ▸The feature implements an open industry specification (ARD) developed collaboratively with Google, GoDaddy, Hugging Face, and Microsoft, promoting ecosystem-wide interoperability
- ▸Enterprises retain full control over which resources agents can access, with no silent auto-installation and governance enforced through the same settings that control Copilot
Summary
GitHub has announced Agent Finder, a new feature for Copilot that enables AI agents to automatically discover and load the right tools, MCP servers, and capabilities based on task descriptions. Rather than pre-configuring all resources upfront, agents can now search an index of available resources in real-time and use only what they need for a specific task.
The feature implements the open Agentic Resource Discovery (ARD) specification, developed collaboratively with Google, GoDaddy, Hugging Face, and Microsoft. Agent Finder searches a registry of AI resources—either GitHub's curated public catalog or a custom private registry—and returns ranked matches that agents can dynamically pull in on demand, helping users avoid manually wiring everything together or burdening their context window.
GitHub emphasizes user control throughout the discovery process: agents cannot auto-install resources, and IT administrators can enforce policies defining which resources agents are permitted to discover and use. Agent Finder is available across all GitHub Copilot plans and works with any registry or AI client that adopts the ARD specification.
Editorial Opinion
Agent Finder represents a meaningful shift toward more flexible and intelligent agent orchestration. By implementing an open specification rather than a proprietary solution, GitHub is positioning interoperability and shared standards as core principles—a smart bet on ecosystem maturity. If other platforms adopt ARD broadly, this could accelerate the industry's move from static tool configuration to dynamic, task-driven capability discovery.
