Navigate Pulse profiling results by scientific setup

.github/workflows/dashboard_refresh.yml

@@ -67,13 +67,20 @@ jobs:
           repository: ${{ github.repository_owner }}/PyAutoMind
           path: _mind
 
+      - name: PyAutoBrain (shared dashboard theme)
+        uses: actions/checkout@v4
+        with:
+          repository: ${{ github.repository_owner }}/PyAutoBrain
+          path: _brain
+
       - name: Dependencies
         run: python3 -m pip install --quiet pyyaml
 
       - name: Dry run (pull request)
         if: github.event_name == 'pull_request'
         env:
           GH_TOKEN: ${{ github.token }}
+          PYAUTO_BRAIN: ${{ github.workspace }}/_brain
         run: |
           bin/pyauto-pulse board --mind _mind
           git status --short
@@ -82,6 +89,7 @@ jobs:
         if: github.event_name != 'pull_request'
         env:
           GH_TOKEN: ${{ github.token }}
+          PYAUTO_BRAIN: ${{ github.workspace }}/_brain
         run: |
           git config user.name "github-actions[bot]"
           git config user.email "41898282+github-actions[bot]@users.noreply.github.com"

.github/workflows/lint.yml

@@ -10,7 +10,7 @@
 #      built from those receipts, and state.json meets the Brain contract.
 #      Offline on purpose: a project commit landing mid-PR must not redden an
 #      unrelated organ PR (dashboard_refresh.yml re-reads the network nightly).
-#   4. lychee link-rot over the prose *.md. raw.githubusercontent.com,
+#   4. lychee link-rot over repository-owned prose *.md (tracked files only). raw.githubusercontent.com,
 #      github.com blob pages (GitHub answers non-browser page views with 503;
 #      a token does not help — the links stay for humans, only the gate skips them),
 #      dashboard.md (hundreds of commit-pinned evidence links), the organ's own
@@ -31,7 +31,7 @@ permissions:
 jobs:
   lint:
     runs-on: ubuntu-latest
-    timeout-minutes: 10
+    timeout-minutes: 15
     steps:
       - uses: actions/checkout@v4
 
@@ -69,6 +69,18 @@ jobs:
           PYAUTO_BRAIN: ${{ github.workspace }}/_brain
         run: python -m pytest tests -q
 
+      - uses: actions/setup-node@v4
+        with:
+          node-version: '22'
+
+      - name: Chromium interaction checks
+        env:
+          PYAUTO_BRAIN: ${{ github.workspace }}/_brain
+        run: |
+          npm install --no-save --no-package-lock playwright@1.63.0
+          npx playwright install --with-deps chromium
+          node tests/browser_setup.cjs
+
       - name: pyauto-pulse check --offline
         env:
           PYAUTO_BRAIN: ${{ github.workspace }}/_brain
@@ -80,6 +92,11 @@ jobs:
           # with a token lychee checks github.com links through the API instead.
           GITHUB_TOKEN: ${{ secrets.GITHUB_TOKEN }}
         run: |
+          # Dependency installs and sibling checkouts are outside this repo's prose.
+          # Keep filenames intact, including spaces, when passing tracked Markdown.
+          mapfile -d '' -t markdown_files < <(
+            git ls-files -z -- '*.md' ':!:dashboard.md' ':!:.claude/**' ':!:_mind/**' ':!:_brain/**'
+          )
           lychee \
             --github-token "${GITHUB_TOKEN}" \
             --exclude '^https://raw\.githubusercontent\.com/' \
@@ -89,4 +106,4 @@ jobs:
             --no-progress \
             --accept '200..=299,429' \
             --max-redirects 5 \
-            $(find . -name '*.md' -not -name 'dashboard.md' -not -path './.claude/*' -not -path './.git/*' -not -path './_mind/*' -not -path './_brain/*')
+            "${markdown_files[@]}"

.gitignore

@@ -17,3 +17,7 @@ _mind/
 # Half-written receipt/snapshot files (written then renamed into place)
 receipts/*.tmp
 snapshots/*.tmp
+
+# Shared theme checkout and browser test dependencies
+_brain/
+node_modules/

AGENTS.md

@@ -121,10 +121,8 @@ catalogue of repositories.
 
 The reader accepts `profiling-summary` v1 and v2; registry rows still pin one
 exact version. V2 adds setup identities, typed evidence and explicit reference
-selections (`pulse/catalogue.py`, `REFERENCE.md`). Keep the live registry on v1
-until the project exporter migration. Accepting a schema is not scientific
-acceptance and does not launch a baseline campaign. The setup browser follows
-in its own phase.
+selections (`pulse/catalogue.py`, `REFERENCE.md`). The live lens registry now reads the project v2 catalogue. Accepting a schema is not scientific
+acceptance and does not launch a baseline campaign. The setup browser reads the captured index and verifies same-commit detail shards.
 
 ## Boundary (what this organ never does)
 

README.md

@@ -18,8 +18,8 @@ It owns:
 - the **instance registry** (`registry.yaml`) — which projects publish a
   profiling summary, where, under which contract version; each project is named
   by its PyAutoMind body-map identity, never a second repository catalogue;
-- the versioned **`profiling-summary` read contract** — v1 is live at
-  `autolens_profiling/dashboard/summary.json`, documented in that repo's
+- the versioned **`profiling-summary` read contract** — v2 is live at
+  `autolens_profiling/dashboard/catalogue.json`, documented in that repo's
   [dashboard/README.md](https://github.com/PyAutoLabs/autolens_profiling/blob/main/dashboard/README.md);
 - the **ingest receipts** (`receipts/`) — the one commit each project was read
   at, per render — and the last-good **snapshots** (`snapshots/`), shown as
@@ -54,3 +54,31 @@ Boundaries: [AGENTS.md](AGENTS.md). Design:
 ## License
 
 MIT — see [LICENSE](LICENSE).
+
+## Setup browser
+
+The front page uses the shared Heart-family theme from `PyAutoBrain/board/_theme.py`.
+Set `PYAUTO_BRAIN` to that checkout (or place it beside this repo / in `_brain`).
+The refresh workflow checks it out explicitly. Open AutoLens, choose a dataset and
+model, then select the instrument and recorded configuration. Runtime is expanded;
+breakdown, compilation, memory, settings, qualification and source evidence are
+available underneath. Values use linear bars within matching measurement groups.
+Campaign tasks live inside their campaign rows; each row has separate evidence
+and task links. Both top actions copy the current editable prompt.
+
+Pulse embeds the validated captured v2 index and loads a chosen evidence shard
+from that **same captured commit**, verifying its SHA-256 and setup identity.
+A failed or cached fetch remains visible; a missing deep link never substitutes a
+different run. Local reads cannot load remote shards as if they were published.
+Inline v2 records and v1 diagnostics remain supported. No rendering action accepts
+a baseline, changes campaign state, runs profiling, or selects the fastest run.
+Temporal diagnostics remain collapsed while the front page focuses on setup
+choices. Existing archive evidence still requires scientific review.
+
+Browser regression checks (synthetic evidence; no project network dependency):
+
+```bash
+npm install --no-save --no-package-lock playwright@1.63.0
+npx playwright install chromium
+node tests/browser_setup.cjs
+```

badge.json

@@ -1,6 +1,6 @@
 {
   "schemaVersion": 1,
   "label": "pulse",
-  "message": "1 project \u00b7 159 records \u00b7 0 cached",
+  "message": "1 project \u00b7 98 records \u00b7 0 cached",
   "color": "yellow"
 }