v0.9.3: home page alignment, confluence search, dead feature flag removals, library additions

[waleedlatif1]

• fix(search): align landing layout and history rows with Home (fix(search): align landing layout and history rows with Home #8317)
• fix(tabs): preserve readable labels in crowded tab strips (fix(tabs): preserve readable labels in crowded tab strips #8319)
• chore(flags): drop leftover references to the retired async projection flag (chore(flags): drop leftover references to the retired async projection flag #8320)
• fix(search): read Confluence pages through the v2 API the Search grant allows (fix(search): read Confluence pages through the v2 API the Search grant allows #8321)
• docs(library): update byok-multi-model-ai-agent-builder (docs(library): update byok-multi-model-ai-agent-builder #8323)
• docs(library): update best-ai-agent-platforms-for-enterprise-teams-2026 (docs(library): update best-ai-agent-platforms-for-enterprise-teams-2026 #8324)

[vercel]

The latest updates on your projects. Learn more about Vercel for GitHub.

1 Skipped Deployment

Project	Deployment	Actions	Updated
docs	Skipped		Sep 26, 2026 5:20pm UTC

[greptile-apps]

Confidence Score: 3/5

[Medium risk] Home page layout refactor and search UI updates.

The PR is not yet safe to merge because edited searches cannot be submitted during an in-flight search and legacy Confluence blog-post references remain unreadable under a selected-space policy.

Findings

1.  Edited searches are cancelled ▶
2.  Legacy blog posts remain unreadable ▶

Summary

The PR aligns organization Home and Search landing layouts, revises search history and source rows, adds a stop control for live search, moves Confluence reads to v2, adjusts crowded tabs, and updates library articles. The new stop control interrupts submission of an edited query while an earlier search is running.

Diagram

%%{init: {'theme': 'neutral'}}%%
flowchart LR
  A[Committed search running] --> B[User edits draft]
  B --> C[Clicks action button]
  C --> D[Current search cancelled]
  D --> E[Edited draft not submitted]

Loading

_{Reviews (2) · Last reviewed commit: "docs(library): update best-ai-agent-plat..."}

[greptile-apps]

Legacy blog posts remain unreadable

When someone opens a saved Confluence blog-post reference from before document kinds were recorded, this scope check requests /pages/{id} and stops on the 404. It runs before the document reader, so the reader’s new fallback to /blogposts/{id} never runs. Under a selected-space policy, the saved blog post cannot be opened.

Knowledge Base Used: Knowledge ingestion, access, and search

[cubic-dev-ai]

1 issue found across 22 files

Confidence score: 3/5

• In apps/sim/lib/sim-search/live/policy.ts, checking only the page endpoint can reject supported kind-less legacy blog-post references before session.read, blocking access; preserve the 404 fallback to the blog-post endpoint.

Prompt for AI agents (unresolved issues)

Check if these issues are valid — if so, understand the root cause of each and fix them. If appropriate, use sub-agents to investigate and fix each issue separately.


<file name="apps/sim/lib/sim-search/live/policy.ts">

<violation number="1" location="apps/sim/lib/sim-search/live/policy.ts:210">
P2: Kind-less live references are supported, but this check probes only the page endpoint, so legacy blog-post references fail authorization before `session.read`. Preserve the 404 fallback to the blog-post endpoint used by `readConfluence`.</violation>
</file>

<sub>Reply with feedback, questions, or to request a fix.

Fix all with cubic | Re-trigger cubic</sub>

[cubic-dev-ai]

P2: Kind-less live references are supported, but this check probes only the page endpoint, so legacy blog-post references fail authorization before session.read. Preserve the 404 fallback to the blog-post endpoint used by readConfluence.

Prompt for AI agents

Check if this issue is valid — if so, understand the root cause and fix it. At apps/sim/lib/sim-search/live/policy.ts, line 210:

<comment>Kind-less live references are supported, but this check probes only the page endpoint, so legacy blog-post references fail authorization before `session.read`. Preserve the 404 fallback to the blog-post endpoint used by `readConfluence`.</comment>

<file context>
@@ -200,14 +200,20 @@ export function createPolicyVerifier(
+      if (!spaceKey) {
+        const row = object(
+          await json(
+            `${api}/${document.kind === 'blogpost' ? 'blogposts' : 'pages'}/${segment(document.id)}`
+          )
         )
</file context>

[greptile-apps]

Edited searches are cancelled

When a user edits the query while the current search is still running, the button now cancels that search instead of submitting the revised text. The user must click again after cancellation to run the new search; previously, they could submit it immediately.

Knowledge Base Used: Web application and product surfaces

Note: If this suggestion doesn't match your team's coding style, reply to this and let me know. I'll remember it for next time!

[cubic-dev-ai]

1 existing issue remains and no new issues found across 22 files

Confidence score: 3/5

• In apps/sim/lib/sim-search/live/policy.ts, references without document.kind are checked only as pages, so supported legacy blogpost references can fail policy verification despite being readable; add the blogpost fallback or reuse readConfluence’s retry behavior.

<sub>Requires human review: Auto-approval blocked because this review re-detected 1 unresolved issue already reported by Cubic.

Re-trigger cubic</sub>