Dashboard FAQ

What is the dashboard?

The dashboard is a living-docs tool for a registered corpus. It runs a three-stage pipeline: refresh benchmarks the corpus and emits a JSON snapshot, render packages that snapshot into a self-contained HTML report, and show pretty-prints the snapshot in your terminal using Rich. Each stage has its own entry point under attune-rag dashboard.

When should I use it?

Use the dashboard when you want to check the health or freshness of a registered corpus — for example, to see how well your RAG pipeline is answering benchmark queries, or to share a snapshot report with teammates as HTML. If you only need to query or update the corpus itself, look at the other features listed in .help/features.yaml.

Which function should I call first?

That depends on what you want to produce:

Each function's signature and return type are documented in the source files listed below.

What happens if queries.yaml is missing?

build_snapshot() returns a partial snapshot dict that includes an error description rather than raising an exception. Check for an error key in the returned dict before passing the snapshot to render() or display().

How do I debug the dashboard?

Start by running the relevant tests:

pytest -k "dashboard" -v

If the tests pass but your code still fails, add a logger.debug statement at the point where the snapshot is built or rendered, then re-run with logging enabled. For known failure modes, see the troubleshooting page for this feature.

Where are the source files?

File Responsibility
src/attune_rag/dashboard/__init__.py Package init
src/attune_rag/dashboard/refresh.py build_snapshot(), main()
src/attune_rag/dashboard/render.py render()
src/attune_rag/dashboard/show.py display()

Tags: dashboard, living-docs, html, terminal, snapshot, freshness