Feature Ideas

We want the catalog to reflect real adoption. Context Hub will begin tracking installs and downloads for skills and MCPs, show useful usage signals in the product, and explore lightweight engagement features such as likes, upvotes, and popularity-based sorting.

We’re connecting support and product feedback to the work that follows. You’ll be able to reach support and share feedback from Context Hub, while Teams requests can become traceable Jira work with their original context preserved. The Frill feedback board and roadmap will make it easier to see what’s being considered and what’s underway.

We’re making quality checks part of the submission journey. Context Hub will run automated checks on submitted skills, look for duplicates, flag security and prompt-injection risks, and show the results directly in the review flow. Approvals will require passing checks, with accountable overrides when an exception is needed.

Submitting a skill should make collaboration easy, not leave feedback scattered. We’re adding conversations between submitters and reviewers, clearer unresolved feedback, visible progress for longer reviews, and in-app and email notifications so everyone knows when a submission needs attention.

We’re expanding Context Hub beyond skills so you can discover MCP servers directly in the catalog and install them through ctx. This also brings better command-line search and information, plus groundwork for reusable bundles or plugins that make it easier to adopt complete AI setups and agent workflows.

Context Hub is the new home for discovering and using approved AI capabilities. We’ve renamed Conduit across the product, documentation, and command-line experience, with the public ctx command and a transition path that keeps existing users supported while they move over.