Add GUI view command to demo-update skill and tensor_accelerator README

demo-update step 0: show user how to open the GUI for the most recent
stage via `bazelisk run //<project>:<module>_<stage> gui_<stage>`.

tensor_accelerator README: add View section, fix build commands to use
bazelisk and target CTS (current most recent stage).

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>

Author: oharboe

.claude/commands/demo-update.md

@@ -4,6 +4,21 @@ Update statistics and screenshots for a demo project.
 
 The user will specify a project name (e.g., `vlsiffra`). Follow these steps:
 
+## 0. View the Result
+
+Ensure the project README has a quick-start GUI command **at the very top**,
+right after the title and description. This lets anyone view the latest
+results immediately:
+
+````markdown
+```bash
+bazelisk run //<project>:<top_module>_<stage> gui_<stage>
+```
+````
+
+Update `<stage>` to the most recent completed stage (e.g., `cts`, `route`, `final`).
+Available stages: `floorplan`, `place`, `cts`, `grt`, `route`, `final`.
+
 ## 1. Run the Flow
 
 Use `/demo-debug <project>` to build incrementally through each stage (_synth → _route → _final),

tensor_accelerator/README.md

@@ -1,5 +1,9 @@
 # Tensor Accelerator — 4-Level Hierarchical TPU
 
+```bash
+bazelisk run //tensor_accelerator:tensor_accelerator_top_cts gui_cts
+```
+
 [tensor_accelerator](https://github.com/profitmonk/tensor_accelerator) by profitmonk
 is an FPGA tensor processing unit with a 4-level macro hierarchy, designed for
 INT8 matrix multiply workloads.
@@ -50,16 +54,16 @@ its ASAP7 performance through CTS.
 
 ```bash
 # Level 0: MAC PE (leaf macro)
-bazel build //tensor_accelerator:mac_pe_place
+bazelisk build //tensor_accelerator:mac_pe_place
 
 # Level 1: Systolic array (mac_pe macros)
-bazel build //tensor_accelerator:systolic_array_generate_abstract
+bazelisk build //tensor_accelerator:systolic_array_generate_abstract
 
 # Level 2: Tensor Processing Cluster (systolic_array macro)
-bazel build //tensor_accelerator:tensor_processing_cluster_generate_abstract
+bazelisk build //tensor_accelerator:tensor_processing_cluster_generate_abstract
 
-# Level 3: Top (4× TPC macros) — through placement
-bazel build //tensor_accelerator:tensor_accelerator_top_place
+# Level 3: Top (4× TPC macros) — through CTS
+bazelisk build //tensor_accelerator:tensor_accelerator_top_cts
 ```
 
 ## References