Feature/pr acceptance criteria

.github/workflows/probe-comment.yml

@@ -10,7 +10,7 @@ jobs:
     name: Post PR Comment
     if: >
       github.event.workflow_run.event == 'pull_request' &&
-      github.event.workflow_run.conclusion == 'success'
+      github.event.workflow_run.conclusion != 'cancelled'
     runs-on: ubuntu-latest
 
     permissions:

.github/workflows/probe.yml

@@ -199,6 +199,18 @@ jobs:
             print('::warning::No probe results found — nothing to report')
             sys.exit(0)
 
+        # ── Baseline gate ─────────────────────────────────────────────
+        BASELINE_TESTS = {'COMP-BASELINE', 'COMP-POST-CL-BODY'}
+        baseline_failures = []
+        for sv in server_data:
+            for r in sv['results']:
+                if r['id'] in BASELINE_TESTS and r['verdict'] == 'Fail':
+                    baseline_failures.append(f"{sv['name']}: {r['id']} — got {r['got']}")
+
+        if baseline_failures:
+            for f in baseline_failures:
+                print(f'::error::{f}')
+
         # ── Write data.js ────────────────────────────────────────────
         output = {
             'commit': {'id': commit_id, 'message': commit_msg, 'timestamp': commit_time},
@@ -238,40 +250,50 @@ jobs:
         def short(tid):
             return re.sub(r'^(RFC\d+-[\d.]+-|COMP-|SMUG-|MAL-|NORM-)', '', tid)
 
+        # Baseline status
+        if baseline_failures:
+            lines.append('### ❌ Baseline Failed')
+            lines.append('')
+            for bf in baseline_failures:
+                lines.append(f'- `{bf}`')
+            lines.append('')
+        else:
+            lines.append('### ✅ Baseline Passed')
+            lines.append('')
+
         for cat_name, title in [('Compliance', 'Compliance'), ('Smuggling', 'Smuggling'), ('MalformedInput', 'Malformed Input'), ('Normalization', 'Header Normalization')]:
             cat_tests = [tid for tid in test_ids if lookup[names[0]][tid]['category'] == cat_name]
             if not cat_tests:
                 continue
             lines.append(f'### {title}')
             lines.append('')
-            # Header row: Server | test1 | test2 | ...
-            hdr = '| Server | ' + ' | '.join(f'`{short(tid)}`' for tid in cat_tests) + ' |'
-            sep = '|---' + ''.join('|:---:' for _ in cat_tests) + '|'
+            # Header row: Test | Expected | Server1 | Server2 | ...
+            hdr = '| Test | Expected | ' + ' | '.join(f'**{n}**' for n in names) + ' |'
+            sep = '|---|---' + ''.join('|:---:' for _ in names) + '|'
             lines.append(hdr)
             lines.append(sep)
-            # Expected row
-            exp_cells = []
+            # One row per test
             for tid in cat_tests:
                 first = lookup[names[0]][tid]
-                exp_cells.append(first['expected'])
-            lines.append('| **Expected** | ' + ' | '.join(exp_cells) + ' |')
-            # Server rows
-            for n in names:
+                expected = first['expected']
                 cells = []
-                for tid in cat_tests:
+                for n in names:
                     r = lookup[n].get(tid)
                     if not r:
                         cells.append('—')
                     else:
                         icon = '✅' if r['verdict'] == 'Pass' else ('⚠️' if r['verdict'] == 'Warn' else '❌')
                         cells.append(f"{icon}`{r['got']}`")
-                lines.append(f"| **{n}** | " + ' | '.join(cells) + ' |')
+                lines.append(f"| `{short(tid)}` | {expected} | " + ' | '.join(cells) + ' |')
             lines.append('')
 
         lines.append(f"<sub>Commit: {commit_id[:7]}</sub>")
 
         with open('probe-comment.md', 'w') as f:
             f.write('\n'.join(lines))
+
+        if baseline_failures:
+            sys.exit(1)
         PYEOF
 
     # ── Upload / publish ───────────────────────────────────────────
@@ -288,7 +310,7 @@ jobs:
       run: echo '${{ github.event.number }}' > pr-number.txt
 
     - name: Upload PR comment
-      if: github.event_name == 'pull_request' && steps.changes.outputs.servers != '[]'
+      if: always() && github.event_name == 'pull_request' && steps.changes.outputs.servers != '[]'
       uses: actions/upload-artifact@v4
       with:
         name: probe-pr-comment