Task-to-skill discovery
Find relevant public skills and inspect their previews and sources through the existing API and MCP integration.
Next proof: useful results and honest previews in actual ChatGPT and Codex conversations.
SkillDB Labs · Active development
Try the next set of SkillDB tools. Each has a useful starting point and a clear next test. Preview status means we are still validating the experience.
Find relevant public skills and inspect their previews and sources through the existing API and MCP integration.
Next proof: useful results and honest previews in actual ChatGPT and Codex conversations.
Small selections for a React app, API service, accessibility review, and data workflow. Each skill has a distinct role.
Next proof: improve representative tasks with limited overlap. Downloads contain references; installation is a separate choice.
Keep workspace-owned practices in a shared library with explicit membership and roles.
Next proof: multi-account access and revocation in a live pilot. Private content stays behind account and workspace checks.
Inspect a draft with explainable checks, define test cases, and keep reproducible records of manual evaluations.
Next proof: repeatable outcomes across real tasks. There is no universal quality score.
Save a public shortlist, compare stated coverage, and copy references into your conversation.
Next proof: authenticated use across devices. Saving is explicit; browser bookmarks are imported only when requested.
A small maintained set of focused workflows, with versions, owners, source references, and evaluation plans.
Next proof: rights review, host behavior, and distribution approval before publication.