docs: the Rugnux card and the naming table mention refinement and live collections

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_015CeEGgpKdfkCTJb3YNMJcu
(cherry picked from commit f8d07397288137dd5e3a0e5c101f2036d51228bd)

docs/NAMING.md

@@ -7,7 +7,7 @@ steep mountain as efficiently as possible — usually by train. Throughput, in o
 | Part | Name | What it does |
 | --- | --- | --- |
 | Streaming / acquisition | **Jungfraujoch** | Receives detector data at high data rates, runs the FPGA/GPU pipeline, and streams images out for writing. |
-| Data processing | **Rugnux** | Offline crystallographic analysis of a stored dataset — indexing, integration, scaling and merging (the [`rugnux`](RUGNUX.md) tool). |
+| Data processing | **Rugnux** | Crystallographic analysis of a stored dataset, or of a collection while it is written — indexing, integration, scaling and merging, and refinement of a given model (the [`rugnux`](RUGNUX.md) tool). |
 
 ## Jungfraujoch
 

docs/documentation.rst

@@ -16,7 +16,8 @@ Documentation
    <p class="jfj-card__text">Processes a stored macromolecular diffraction dataset, a rotation
    sweep or serial stills, from the images to merged reflections: spot finding, indexing,
    integration, scaling and merging, with the space group, cell and resolution decided from the
-   data and a report that says on what evidence. One command, no parameters.</p>
+   data and a report that says on what evidence; given an atomic model, it also places and refines
+   it against the merge. One command, no parameters.</p>
    <pre class="jfj-card__run"><code>rugnux &lt;diffraction image file&gt;</code></pre>
    <p class="jfj-platforms">Linux &middot; macOS &middot; Windows &middot; fast on an NVIDIA GPU,
    works without one</p>