Premium skill

Ground your AI SDK agent in real data

Measured retrieval for AI SDK agents: corpus-derived chunking, hybrid search, citations in the UI, and golden-set evals. Prove it works before you ship it.

Install with Premium View pricing

Plan: Premium
Access: Gated
Install: Tokenized
Authored files: 8

Install with skills CLIPremium skills install through a short-lived private link.

$ npx skills add … --skill ground-ai-sdk-agent

Open Premium install

Filesystem

Browse the authored file tree — tools, references, and evals included. Source unlocks with Premium.

SKILL.mdmarkdown

Unlock skill source

Browse every authored SKILL.md with premium.

Install with Premium

File tree for ground-ai-sdk-agent stays visible so you can see the pack before unlocking.

File

What you get

Corpus first, chunking second

Audit formats, structure, and access rules before choosing a chunker, store, or embedding model.

Citations as typed parts

Emit source or data parts from the server and wire them into the chat surface. No markdown scraping.

Measured retrieval

Golden-set recall@k and MRR before and after every ingest or ranking change. Live graders stay opt-in.

ACL in the query

Tenant and role filters run in retrieval code. Cross-tenant leakage fails a deterministic eval.

Ship RAG you can defend

The skill pairs with Better Chat UI for citation surfaces and Build AI SDK Agent Evals for dataset and scorer conventions.

A docs chatbot that cites the exact section it used.
A support agent that abstains when the knowledge base has no answer.
A multi-tenant policy search that cannot leak across companies.
An existing RAG stack that answers confidently wrong: diagnose retrieval vs generation with metrics.

Stop shipping vibed retrieval

Available with Premium. The private install link expires; the downloaded skill does not. Generate a fresh link when you want to update it.