feat(crunchbase): add Crunchbase Data API integration (#6875)

* feat(crunchbase): add Crunchbase Data API integration

Covers the v4 Data API end to end: dedicated search and lookup operations
for organizations, people, funding rounds, and acquisitions, plus generic
collection-parameterized search and lookup reaching the remaining 39
collections, single-card paging, autocomplete, the deleted-entity feed, and
fields metadata.

Adds a crunchbase-errors extractor: the API answers failures with a bare
JSON array, which no existing extractor reads, so an auth or predicate
failure would have reported only its HTTP status.

* fix(crunchbase): honor card paging limits and cursor exclusivity

- Cap a card page at the documented 100-item maximum instead of Search's
  1000, which the shared Limit field made easy to carry over
- Always request the card's identifier so a narrowed cardFieldIds cannot
  return a full page with a null cursor and stall a paging loop
- Reject the mutually-exclusive afterId/beforeId pair on the card and
  deleted-entity endpoints, not just on search
- Report an unexpected card shape as empty rather than wrapping the
  envelope as a one-row page

Author: waleedlatif1

apps/docs/components/icons.tsx

@@ -1489,15 +1489,15 @@ export function InstagramIcon(props: SVGProps<SVGSVGElement>) {
 
 export function CrunchbaseIcon(props: SVGProps<SVGSVGElement>) {
   return (
-    <svg
-      {...props}
-      fill='currentColor'
-      width='24'
-      height='24'
-      viewBox='0 0 24 24'
-      xmlns='http://www.w3.org/2000/svg'
-    >
-      <path d='M21.6 0H2.4A2.41 2.41 0 0 0 0 2.4v19.2A2.41 2.41 0 0 0 2.4 24h19.2a2.41 2.41 0 0 0 2.4-2.4V2.4A2.41 2.41 0 0 0 21.6 0zM7.05 14.47A2.11 2.11 0 0 0 9.84 13.42h1.66a3.69 3.69 0 1 1 0-1.75H9.84a2.11 2.11 0 1 0-2.79 2.8zm11.35.845a3.55 3.55 0 0 1-1.06.63 3.68 3.68 0 0 1-3.39-.38v.38h-1.51V5.37h1.5v4.11a3.74 3.74 0 0 1 1.8-.63H16a3.67 3.67 0 0 1 2.39 6.46zm-.223-2.77a2.1 2.1 0 1 1-4.21 0 2.1 2.1 0 0 1 4.21 0z' />
+    <svg {...props} viewBox='0 0 32 32' fill='none' xmlns='http://www.w3.org/2000/svg'>
+      <path
+        d='M28.802 0h-25.604c-1.76 0.005-3.193 1.438-3.198 3.198v25.604c0.005 1.76 1.438 3.193 3.198 3.198h25.604c1.76-0.005 3.193-1.438 3.198-3.198v-25.604c-0.005-1.76-1.438-3.193-3.198-3.198z'
+        fill='#ffffff'
+      />
+      <path
+        d='M28.802 0h-25.604c-1.76 0.005-3.193 1.438-3.198 3.198v25.604c0.005 1.76 1.438 3.193 3.198 3.198h25.604c1.76-0.005 3.193-1.438 3.198-3.198v-25.604c-0.005-1.76-1.438-3.193-3.198-3.198zM9.396 19.286c1.411 0.646 3.078 0.021 3.724-1.391h2.214c-1.38 5.651-9.698 4.651-9.698-1.167 0-5.823 8.318-6.823 9.698-1.167h-2.214c-0.813-1.786-3.161-2.214-4.547-0.823-1.391 1.385-0.964 3.734 0.823 4.547zM24.521 20.411c-0.422 0.365-0.896 0.646-1.417 0.844-1.495 0.578-3.182 0.391-4.516-0.51v0.51h-2.016v-14.094h2v5.479c0.714-0.484 1.542-0.771 2.401-0.839h0.359c4.552-0.010 6.646 5.656 3.188 8.609zM24.224 16.724c0.031 1.573-1.234 2.87-2.807 2.87s-2.839-1.297-2.802-2.87c0.078-3.656 5.526-3.656 5.609 0z'
+        fill='#0287d1'
+      />
     </svg>
   )
 }

apps/docs/components/ui/icon-mapping.ts

@@ -47,6 +47,7 @@ import {
   ContextDevIcon,
   ConvexIcon,
   CrowdStrikeIcon,
+  CrunchbaseIcon,
   CursorIcon,
   DagsterIcon,
   DatabricksIcon,
@@ -307,6 +308,7 @@ export const blockTypeToIconMap: Record<string, IconComponent> = {
   context_dev: ContextDevIcon,
   convex: ConvexIcon,
   crowdstrike: CrowdStrikeIcon,
+  crunchbase: CrunchbaseIcon,
   cursor: CursorIcon,
   cursor_v2: CursorIcon,
   dagster: DagsterIcon,

apps/docs/content/docs/en/integrations/crunchbase.mdx

@@ -48,6 +48,7 @@
     "context_dev",
     "convex",
     "crowdstrike",
+    "crunchbase",
     "cursor",
     "dagster",
     "databricks",

apps/docs/content/docs/en/integrations/meta.json

@@ -0,0 +1,175 @@
+/**
+ * @vitest-environment node
+ */
+import { describe, expect, it } from 'vitest'
+import { CrunchbaseBlock } from '@/blocks/blocks/crunchbase'
+
+/**
+ * Every assertion here runs against `{ ...inputs, ...buildParams(inputs) }`, the
+ * shape the generic tool handler actually forwards. A key the mapper omits is
+ * *not* dropped by that merge — the raw subBlock value survives — so asserting
+ * on the mapper's return alone would prove nothing about what the tool receives.
+ */
+describe('CrunchbaseBlock', () => {
+  const buildParams = CrunchbaseBlock.tools.config!.params!
+  const selectTool = CrunchbaseBlock.tools.config!.tool!
+
+  const resolve = (inputs: Record<string, unknown>) => ({ ...inputs, ...buildParams(inputs) })
+
+  const operationIds =
+    CrunchbaseBlock.subBlocks
+      .find((subBlock) => subBlock.id === 'operation')
+      ?.options?.map((option) => (option as { id: string }).id) ?? []
+
+  it('maps every dropdown operation onto a registered tool', () => {
+    expect(operationIds).toHaveLength(14)
+    expect(new Set(operationIds.map((id) => selectTool({ operation: id })))).toEqual(
+      new Set(CrunchbaseBlock.tools.access)
+    )
+  })
+
+  it('rejects an operation the dropdown does not offer', () => {
+    expect(() => selectTool({ operation: 'search_unicorns' })).toThrow(
+      /Invalid Crunchbase operation/
+    )
+  })
+
+  it('gives every subblock a unique id', () => {
+    const ids = CrunchbaseBlock.subBlocks.map((subBlock) => subBlock.id)
+    expect(ids).toHaveLength(new Set(ids).size)
+  })
+
+  it('shows a subblock for every operation that owns it', () => {
+    const conditionsFor = (id: string) => {
+      const condition = CrunchbaseBlock.subBlocks.find((subBlock) => subBlock.id === id)?.condition
+      const value = (condition as { value?: unknown } | undefined)?.value
+      return new Set(Array.isArray(value) ? value.map(String) : [String(value)])
+    }
+
+    expect(conditionsFor('searchQuery')).toEqual(
+      new Set([
+        'search_organizations',
+        'search_people',
+        'search_funding_rounds',
+        'search_acquisitions',
+        'search_entities',
+      ])
+    )
+    expect(conditionsFor('entityId')).toEqual(
+      new Set([
+        'get_organization',
+        'get_person',
+        'get_funding_round',
+        'get_acquisition',
+        'get_entity',
+        'get_entity_card',
+      ])
+    )
+  })
+
+  it('never hides a required field behind advanced mode', () => {
+    const advancedRequired = CrunchbaseBlock.subBlocks
+      .filter((subBlock) => subBlock.mode === 'advanced' && subBlock.required)
+      .map((subBlock) => subBlock.id)
+
+    expect(advancedRequired).toEqual([])
+  })
+
+  /*
+   * The mapper's whole job is dropping keys that belong to another operation.
+   * `shouldSerializeSubBlock` short-circuits on `mode: 'advanced'` before it
+   * evaluates a condition, so a hidden advanced field still reaches `inputs` —
+   * only an explicit `undefined` removes it from what goes out on the wire.
+   */
+  it('drops a previous operation’s leftovers when the operation changes', () => {
+    const params = resolve({
+      operation: 'autocomplete',
+      apiKey: 'key',
+      autocompleteQuery: 'airbnb',
+      searchQuery: '[{"type":"predicate","field_id":"categories"}]',
+      searchFieldIds: '["identifier"]',
+      entityId: 'tesla-motors',
+      fieldIds: '["identifier","name"]',
+      cardIds: '["founders"]',
+      order: '[{"field_id":"rank_org","sort":"asc"}]',
+      afterId: 'stale-cursor',
+      collection: 'organizations',
+    })
+
+    expect(params.query).toBe('airbnb')
+    expect(params.entityId).toBeUndefined()
+    expect(params.fieldIds).toBeUndefined()
+    expect(params.cardIds).toBeUndefined()
+    expect(params.order).toBeUndefined()
+    expect(params.afterId).toBeUndefined()
+    expect(params.collection).toBeUndefined()
+    expect(params.searchQuery).toBeUndefined()
+    expect(params.autocompleteQuery).toBeUndefined()
+  })
+
+  it('routes the predicate query to a search and the text query to autocomplete', () => {
+    const search = resolve({
+      operation: 'search_organizations',
+      apiKey: 'key',
+      searchQuery: '[{"type":"predicate","field_id":"categories","operator_id":"includes"}]',
+      autocompleteQuery: 'airbnb',
+    })
+    expect(search.query).toBe(
+      '[{"type":"predicate","field_id":"categories","operator_id":"includes"}]'
+    )
+
+    const autocomplete = resolve({
+      operation: 'autocomplete',
+      apiKey: 'key',
+      searchQuery: '[{"type":"predicate","field_id":"categories","operator_id":"includes"}]',
+      autocompleteQuery: 'airbnb',
+    })
+    expect(autocomplete.query).toBe('airbnb')
+  })
+
+  it('feeds the generic search its own required field list', () => {
+    const generic = resolve({
+      operation: 'search_entities',
+      apiKey: 'key',
+      collection: 'events',
+      searchQuery: '[]',
+      searchFieldIds: '["identifier","short_description"]',
+      fieldIds: '["identifier","name"]',
+    })
+    expect(generic.fieldIds).toBe('["identifier","short_description"]')
+    expect(generic.searchFieldIds).toBeUndefined()
+
+    const specific = resolve({
+      operation: 'search_organizations',
+      apiKey: 'key',
+      searchQuery: '[]',
+      searchFieldIds: '["identifier","short_description"]',
+      fieldIds: '["identifier","name"]',
+    })
+    expect(specific.fieldIds).toBe('["identifier","name"]')
+  })
+
+  it('picks the collection each operation actually asks for', () => {
+    const card = resolve({
+      operation: 'get_entity_card',
+      apiKey: 'key',
+      entityId: 'sequoia-capital',
+      cardId: 'participated_investments',
+      cardCollection: 'organizations',
+      collection: 'events',
+      deletedCollection: 'people',
+    })
+    expect(card.collection).toBe('organizations')
+    expect(card.cardCollection).toBeUndefined()
+
+    const deleted = resolve({
+      operation: 'list_deleted_entities',
+      apiKey: 'key',
+      cardCollection: 'organizations',
+      collection: 'events',
+      deletedCollection: 'people',
+    })
+    expect(deleted.collection).toBe('people')
+    expect(deleted.deletedCollection).toBeUndefined()
+  })
+})