fix(search): harden source parsing, availability, and provider matching

- validate <source> urls by parsing them and requiring a host, so a
  malformed value never renders a dead citation link
- gate Search-mode suggestions on deployment OAuth availability, sharing
  the predicate with the Search catalog
- match connected credentials across a service's additional provider ids
  (Salesforce sandbox) and count them in connector telemetry
- collect footer sources from the rendered text segments, covering a
  block-less message's fallback text and excluding subagent lanes
- use a distinctive citation-link sentinel and an absolute import for
  SuggestedActions
- teach the email tokens transcription test the composed chip geometry

Author: emir-karabeg

apps/sim/app/workspace/[workspaceId]/home/components/message-content/components/chat-content/chat-content.tsx

@@ -128,14 +128,19 @@ function nextInlineSegmentLabel(segment?: ContentSegment): string {
 
 /**
  * The `<source>` payloads of the segment being rendered, in emission order. An
- * inline citation is written into the markdown as `[label](#src-N)` so it flows
- * with its paragraph, and the link renderer resolves `N` back through this
- * context — the component map is static, so it is the one channel from segment
- * data into it.
+ * inline citation is written into the markdown as a link to a sentinel
+ * fragment carrying the payload's index, so it flows with its paragraph, and
+ * the link renderer resolves the index back through this context — the
+ * component map is static, so it is the one channel from segment data into it.
  */
 const SourceRefsContext = createContext<readonly SourceTagData[]>([])
 
-const SOURCE_LINK_PREFIX = '#src-'
+/**
+ * Fragment prefix of a generated citation link. Internal — never navigated —
+ * and deliberately not a name the model would write on its own; an index that
+ * resolves to no parsed source falls back to the link text.
+ */
+const SOURCE_LINK_PREFIX = '#sim-source-ref-'
 
 interface SourceReferenceProps {
   index: number

apps/sim/app/workspace/[workspaceId]/home/components/message-content/components/special-tags/special-tags.test.ts

@@ -1407,12 +1407,14 @@ describe('source tag', () => {
   })
 
   it('rejects a source without an absolute http(s) url', () => {
-    const { segments } = parseSpecialTags(
-      'See <source>{"url":"docs/internal.md","siteName":"Docs"}</source>.',
-      false
-    )
+    for (const url of ['docs/internal.md', 'https://?', 'ftp://host/x', 'https://a b.example/x']) {
+      const { segments } = parseSpecialTags(
+        `See <source>{"url":"${url}","siteName":"Docs"}</source>.`,
+        false
+      )
 
-    expect(segments.some((segment) => segment.type === 'source')).toBe(false)
+      expect(segments.some((segment) => segment.type === 'source')).toBe(false)
+    }
   })
 
   it('hides a half-arrived source opener while streaming', () => {

apps/sim/app/workspace/[workspaceId]/home/components/message-content/components/special-tags/special-tags.tsx

@@ -556,9 +556,19 @@ function isMothershipErrorTagData(value: unknown): value is MothershipErrorTagDa
   )
 }
 
-/** Only an absolute http(s) URL can be linked; anything else is not a source. */
+/**
+ * Only an absolute http(s) URL with a host can be linked; anything else is not
+ * a source. Parsed rather than pattern-matched so a malformed value such as
+ * `https://?` — which a prefix check would accept — never becomes a dead link.
+ */
 function isHttpUrl(value: unknown): value is string {
-  return typeof value === 'string' && /^https?:\/\/\S+$/i.test(value.trim())
+  if (typeof value !== 'string' || /\s/.test(value)) return false
+  try {
+    const url = new URL(value)
+    return (url.protocol === 'http:' || url.protocol === 'https:') && url.hostname.length > 0
+  } catch {
+    return false
+  }
 }
 
 function isSourceTagData(value: unknown): value is SourceTagData {

apps/sim/app/workspace/[workspaceId]/home/components/message-content/message-content.tsx

@@ -864,7 +864,6 @@ function MessageContentInner({
     () => (blocks.length > 0 ? parseBlocks(blocks) : []),
     [blocks, blockOverlayVersion]
   )
-  const sources = useMemo(() => collectMessageSources(blocks), [blocks])
 
   const [trailingRevealing, setTrailingRevealing] = useState(false)
   const handleTrailingRevealChange = useCallback((revealing: boolean) => {
@@ -880,12 +879,28 @@ function MessageContentInner({
   }, [])
   const [isStreamIdle, setIsStreamIdle] = useState(false)
 
-  const segments: MessageSegment[] =
-    parsed.length > 0
-      ? parsed
-      : fallbackContent?.trim()
-        ? [{ type: 'text' as const, id: 'text-fallback', content: fallbackContent }]
-        : []
+  const segments = useMemo<MessageSegment[]>(
+    () =>
+      parsed.length > 0
+        ? parsed
+        : fallbackContent?.trim()
+          ? [{ type: 'text', id: 'text-fallback', content: fallbackContent }]
+          : [],
+    [parsed, fallbackContent]
+  )
+  /**
+   * Collected from the segments that render, not the raw blocks: that is the
+   * same text the inline chips come from, so the footer agrees with them — it
+   * covers the fallback text of a block-less message and leaves out lane text
+   * that `parseBlocks` folds into agent groups.
+   */
+  const sources = useMemo(
+    () =>
+      collectMessageSources(
+        segments.flatMap((segment) => (segment.type === 'text' ? [segment.content] : []))
+      ),
+    [segments]
+  )
   const visibleStreamActivityKey = getVisibleStreamActivityKey(segments)
 
   // Every visible stream update restarts the quiet-period clock. A layout

apps/sim/app/workspace/[workspaceId]/home/components/message-content/utils.test.ts

@@ -2,7 +2,6 @@
  * @vitest-environment node
  */
 import { describe, expect, it } from 'vitest'
-import { ContentBlockType } from '@/app/workspace/[workspaceId]/home/types'
 import { collectMessageSources, deriveMessagePhase, resolveToolDisplayState } from './utils'
 
 describe('deriveMessagePhase', () => {
@@ -41,33 +40,21 @@ describe('resolveToolDisplayState', () => {
 describe('collectMessageSources', () => {
   const source = (url: string, extra = '') => `<source>{"url":"${url}"${extra}}</source>`
 
-  it('collects every distinct source across the message text, in first-cited order', () => {
-    const blocks = [
-      {
-        type: ContentBlockType.text,
-        content: `First point. ${source('https://a.example/1', ',"siteName":"A"')} Second. ${source('https://b.example/2')}`,
-      },
-      { type: ContentBlockType.tool_call },
-      {
-        type: ContentBlockType.text,
-        content: `Again. ${source('https://a.example/1')} New. ${source('https://c.example/3')}`,
-      },
+  it('collects every distinct source across the given text, in first-cited order', () => {
+    const texts = [
+      `First point. ${source('https://a.example/1', ',"siteName":"A"')} Second. ${source('https://b.example/2')}`,
+      `Again. ${source('https://a.example/1')} New. ${source('https://c.example/3')}`,
     ]
 
-    expect(collectMessageSources(blocks).map((entry) => entry.url)).toEqual([
+    expect(collectMessageSources(texts).map((entry) => entry.url)).toEqual([
       'https://a.example/1',
       'https://b.example/2',
       'https://c.example/3',
     ])
-    expect(collectMessageSources(blocks)[0].siteName).toBe('A')
+    expect(collectMessageSources(texts)[0].siteName).toBe('A')
   })
 
-  it('ignores subagent lanes and text without sources', () => {
-    const blocks = [
-      { type: ContentBlockType.subagent_text, content: source('https://lane.example/x') },
-      { type: ContentBlockType.text, content: 'Plain prose.' },
-    ]
-
-    expect(collectMessageSources(blocks)).toEqual([])
+  it('returns nothing for prose without sources', () => {
+    expect(collectMessageSources(['Plain prose.', ''])).toEqual([])
   })
 })

apps/sim/app/workspace/[workspaceId]/home/components/message-content/utils.ts

@@ -25,25 +25,19 @@ import {
   parseSpecialTags,
   type SourceTagData,
 } from '@/app/workspace/[workspaceId]/home/components/message-content/components/special-tags'
-import {
-  type ContentBlock,
-  ContentBlockType,
-  type ToolCallStatus,
-} from '@/app/workspace/[workspaceId]/home/types'
+import type { ToolCallStatus } from '@/app/workspace/[workspaceId]/home/types'
 
 export type IconComponent = ComponentType<SVGProps<SVGSVGElement>>
 
 /**
- * Every distinct `<source>` cited in the message's own prose, in first-cited
- * order, for the footer strip. Only main-lane text counts: subagent lanes fold
- * into agent groups rather than the answer, and a tool's output is not a
- * citation.
+ * Every distinct `<source>` cited across the given prose, in first-cited order,
+ * for the footer strip. Callers pass the text segments the message actually
+ * renders as its answer.
  */
-export function collectMessageSources(blocks: ContentBlock[]): SourceTagData[] {
+export function collectMessageSources(texts: readonly string[]): SourceTagData[] {
   const byUrl = new Map<string, SourceTagData>()
-  for (const block of blocks) {
-    if (block.type !== ContentBlockType.text || !block.content) continue
-    for (const segment of parseSpecialTags(block.content, false).segments) {
+  for (const text of texts) {
+    for (const segment of parseSpecialTags(text, false).segments) {
       if (segment.type === 'source' && !byUrl.has(segment.data.url)) {
         byUrl.set(segment.data.url, segment.data)
       }

apps/sim/app/workspace/[workspaceId]/home/components/suggested-actions/connector-actions.test.ts

@@ -15,12 +15,17 @@ vi.mock('@/lib/sim-search/connectors', () => {
     type,
     meta: { id: type, name, description: `Sync ${name}`, icon },
     providerId,
+    providerIds: [providerId],
     requiredScopes: [],
     serviceName: name,
     serviceIcon: icon,
     blockType: type,
   })
   return {
+    isSearchConnectorConnected: (
+      candidate: { providerIds: string[] },
+      connected: ReadonlySet<string>
+    ) => candidate.providerIds.some((providerId) => connected.has(providerId)),
     SEARCH_CONNECTORS: [
       connector('airtable', 'Airtable', 'airtable'),
       connector('confluence', 'Confluence', 'confluence'),
@@ -34,14 +39,16 @@ vi.mock('@/lib/sim-search/connectors', () => {
 
 import { computeConnectorActions } from '@/app/workspace/[workspaceId]/home/components/suggested-actions/connector-actions'
 
+const ALL_AVAILABLE = () => true
+
 describe('computeConnectorActions', () => {
   beforeEach(() => {
     /** A zero roll always samples the first remaining candidate, so the rotation is catalog order. */
     mockRandomFloat.mockReturnValue(0)
   })
 
   it('pins Confluence, Jira, and JSM first and fills the last slot from the rotation', () => {
-    const actions = computeConnectorActions(new Set())
+    const actions = computeConnectorActions(new Set(), ALL_AVAILABLE)
 
     expect(actions.map((action) => action.id)).toEqual([
       'connect-confluence',
@@ -57,17 +64,31 @@ describe('computeConnectorActions', () => {
   })
 
   it('drops every connector on a connected provider and refills from the rotation', () => {
-    const actions = computeConnectorActions(new Set(['jira', 'airtable']))
+    const actions = computeConnectorActions(new Set(['jira', 'airtable']), ALL_AVAILABLE)
+
+    expect(actions.map((action) => action.id)).toEqual([
+      'connect-confluence',
+      'connect-notion',
+      'connect-slack',
+    ])
+  })
+
+  it('drops connectors this deployment cannot connect, pinned or not', () => {
+    const actions = computeConnectorActions(
+      new Set(),
+      (connector) => connector.type !== 'jira' && connector.type !== 'airtable'
+    )
 
     expect(actions.map((action) => action.id)).toEqual([
       'connect-confluence',
+      'connect-jsm',
       'connect-notion',
       'connect-slack',
     ])
   })
 
   it('returns fewer than four rows once the rotation is exhausted', () => {
-    const actions = computeConnectorActions(new Set(['airtable', 'notion', 'slack']))
+    const actions = computeConnectorActions(new Set(['airtable', 'notion', 'slack']), ALL_AVAILABLE)
 
     expect(actions.map((action) => action.id)).toEqual([
       'connect-confluence',

apps/sim/app/workspace/[workspaceId]/home/components/suggested-actions/connector-actions.ts

@@ -1,4 +1,8 @@
-import { SEARCH_CONNECTORS, type SearchConnector } from '@/lib/sim-search/connectors'
+import {
+  isSearchConnectorConnected,
+  SEARCH_CONNECTORS,
+  type SearchConnector,
+} from '@/lib/sim-search/connectors'
 import type { Action } from '@/app/workspace/[workspaceId]/home/components/suggested-actions/types'
 import { weightedSample } from '@/app/workspace/[workspaceId]/home/components/suggested-actions/weighted-sample'
 
@@ -39,12 +43,17 @@ function toConnectorAction(connector: SearchConnector): Action {
  * sample of the rest to fill four slots. A connector whose provider the viewer
  * has already connected is dropped from both halves — so Jira and Jira Service
  * Management, which share one provider, leave together — and a pinned slot
- * freed that way is taken by the rotation.
+ * freed that way is taken by the rotation. Connectors this deployment cannot
+ * connect are dropped the same way, so a row never opens a modal that fails.
  */
-export function computeConnectorActions(connectedProviderIds: ReadonlySet<string>): Action[] {
-  const isConnected = (connector: SearchConnector) => connectedProviderIds.has(connector.providerId)
-  const pinned = PINNED.filter((connector) => !isConnected(connector))
-  const pool = ROTATING.filter((connector) => !isConnected(connector))
+export function computeConnectorActions(
+  connectedProviderIds: ReadonlySet<string>,
+  isAvailable: (connector: SearchConnector) => boolean
+): Action[] {
+  const offered = (connector: SearchConnector) =>
+    isAvailable(connector) && !isSearchConnectorConnected(connector, connectedProviderIds)
+  const pinned = PINNED.filter(offered)
+  const pool = ROTATING.filter(offered)
   const rotating = weightedSample(pool, CONNECTOR_ACTION_COUNT - pinned.length, () => 1)
   return [...pinned, ...rotating].map(toConnectorAction)
 }

apps/sim/app/workspace/[workspaceId]/home/components/suggested-actions/suggested-actions.test.tsx

@@ -32,6 +32,11 @@ vi.mock('@/hooks/queries/kb/knowledge', () => ({
 vi.mock('@/app/workspace/[workspaceId]/search/hooks/use-search-credentials', () => ({
   useSearchCredentials: mockUseSearchCredentials,
 }))
+vi.mock('@/hooks/use-permission-config', () => ({
+  usePermissionConfig: () => ({
+    integrationAvailability: new Map([['notion', { state: 'unavailable', oauthAvailable: false }]]),
+  }),
+}))
 
 /** The Build-mode pool is built from the block catalog at module load; an empty catalog keeps it to the table starters. */
 vi.mock('@/blocks/registry', () => ({ getAllBlockMeta: () => ({}), getAllBlocks: () => [] }))
@@ -42,18 +47,28 @@ vi.mock('@/lib/sim-search/connectors', () => {
     type,
     meta: { id: type, name, description: `Sync ${name}`, icon },
     providerId,
+    providerIds: [providerId],
     requiredScopes: ['read'],
     serviceName: name,
     serviceIcon: icon,
     blockType: type,
   })
   return {
+    isSearchConnectorConnected: (
+      candidate: { providerIds: string[] },
+      connected: ReadonlySet<string>
+    ) => candidate.providerIds.some((providerId) => connected.has(providerId)),
+    isSearchConnectorAvailable: (
+      candidate: { blockType: string },
+      availability: ReadonlyMap<string, { oauthAvailable: boolean }>
+    ) => availability.get(candidate.blockType)?.oauthAvailable ?? true,
     SEARCH_CONNECTORS: [
       connector('airtable', 'Airtable', 'airtable'),
       connector('confluence', 'Confluence', 'confluence'),
       connector('jira', 'Jira', 'jira'),
       connector('jsm', 'Jira Service Management', 'jira'),
       connector('notion', 'Notion', 'notion'),
+      connector('slack', 'Slack', 'slack'),
     ],
   }
 })
@@ -117,7 +132,7 @@ describe('SuggestedActions', () => {
     expect(rows().map((row) => row.textContent)).toContain('Integrate with Slack')
   })
 
-  it('swaps to the connector list in Search mode, minus providers already connected', () => {
+  it('swaps to the connector list in Search mode, minus connected and unavailable connectors', () => {
     mount()
 
     act(() => useMothershipModeStore.getState().setMode('search'))
@@ -126,7 +141,7 @@ describe('SuggestedActions', () => {
     expect(rows().map((row) => row.textContent)).toEqual([
       'Connect Confluence',
       'Connect Airtable',
-      'Connect Notion',
+      'Connect Slack',
     ])
   })
 
@@ -144,7 +159,12 @@ describe('SuggestedActions', () => {
     expect(mockCaptureEvent).toHaveBeenCalledWith(
       null,
       'suggested_action_clicked',
-      expect.objectContaining({ kind: 'connector', action_id: 'connect-confluence', position: 0 })
+      expect.objectContaining({
+        kind: 'connector',
+        action_id: 'connect-confluence',
+        position: 0,
+        connected_provider_count: 1,
+      })
     )
   })
 })