NVIDIA AI-Q — глубокие исследования локально

★ 7.6 · content

aiq-research is a Claude Code skill that runs deep research queries against a locally deployed NVIDIA AI-Q Blueprint server. It uses the helper script `scripts/aiq.py` to check backend availability with a `health` call, submit questions via `chat`, and — when the server returns an asynchronous job — poll for completion using `research_poll`, `status`, and `report` commands. Finished reports are presented with citations and source URLs intact; generated artifacts such as charts and CSVs can be downloaded with the `artifacts` command. Python 3.11+ and a running AI-Q backend are required — defaulting to `http://localhost:8000`, overridable via the `AIQ_SERVER_URL` environment variable. The skill is designed for requests like "deep research on X" or "use AI-Q to answer Y" whenever a sourced, structured report is needed.