A shortlist with a reason behind it.
Choose the task you need to complete and the priority that matters most. The task selects up to three candidates from the curated catalogue; your priority shapes the review guidance. Each candidate comes with a reason to consider it, provider references and a limitation to check.
An optional team note helps you keep your context alongside the result. It stays with your plan and does not change the candidate selection.
From a choice to a useful next step.
- Describe the work. Choose a task and a priority. Add a short team note if it helps explain the decision.
- Review the candidates. Read the fit, open the sources and check the limitations. Use the suggested workflow and edit the starting prompt for your material.
- Keep what you need. Copy the result or export Markdown. Choose Save on this device to keep a shortlist, then manage it in Saved shortlists. You can keep up to eight.
Choose with the evidence in view.
This is curated, criteria-based selection from the Hub’s catalogue. Candidates are starting points for a decision, not measured rankings. Check a provider’s current capabilities, access requirements and terms before using it.
The Hub helps you prepare the work. You open the chosen provider separately, supply material you are allowed to use, and review the output yourself. The workflow guides explain what to prepare and what to check.
What is in development.
A planned Claude API integration would interpret a team’s requirements and explain candidate trade-offs against catalogue sources. It is in development and is not connected to the public selector. The current selector makes no AI API call and does not execute a vendor tool.
Planned evaluation covers source coverage, task fit, unsupported capability claims and whether people can understand and revise suggestions. These are evaluation goals, not measured results or a release commitment.
For a selection question that needs a conversation, contact the project operator.