restructure recommendations for index targeting in md guides, npm run format, fix VERCEL_TOKEN not being threaded through from to the release -> pr flow properly

Author: austin-denoble

.github/workflows/pr.yml

@@ -15,6 +15,8 @@ on:
     secrets:
       PINECONE_API_KEY:
         required: true
+      VERCEL_TOKEN:
+        required: true
 
 permissions:
   contents: read

README.md

@@ -77,7 +77,7 @@ const pc = new Pinecone({ apiKey: 'YOUR_API_KEY' });
 // const pc = new Pinecone();
 
 // 2. Create a serverless index
-await pc.createIndex({
+const indexModel = await pc.createIndex({
   name: 'example-index',
   dimension: 1536,
   metric: 'cosine',
@@ -90,7 +90,7 @@ await pc.createIndex({
 });
 
 // 3. Target the index
-const index = pc.index({ name: 'example-index' });
+const index = pc.index({ host: indexModel.host });
 
 // 4. Upsert vectors with metadata
 await index.upsert({
@@ -193,7 +193,7 @@ console.log(searchResponse);
 
 ## Pinecone Assistant
 
-The [Pinecone Assistant API](https://docs.pinecone.io/guides/assistant/understanding-assistant) enables you to create and manage AI assistants powered by Pinecone's vector database capabilities. These Assistants can be customized with specific instructions and metadata, and can interact with files and engage in chat conversations.
+The [Pinecone Assistant API](https://docs.pinecone.io/guides/assistant) enables you to create and manage AI assistants powered by Pinecone's vector database capabilities. These Assistants can be customized with specific instructions and metadata, and can interact with files and engage in chat conversations.
 
 ```typescript
 import { Pinecone } from '@pinecone-database/pinecone';

guides/data-operations/bulk-import.md

@@ -1,6 +1,6 @@
 # Bulk Import
 
-You can [import vectors en masse](https://docs.pinecone.io/guides/index-data/import-data) from object storage. `Import` is a long-running, asynchronous operation that imports large numbers of records into a Pinecone serverless index.
+You can [import vectors in bulk](https://docs.pinecone.io/guides/index-data/import-data) from object storage. `Import` is a long-running, asynchronous operation that imports large numbers of records into a Pinecone serverless index.
 
 > **Note:**
 >
@@ -24,7 +24,7 @@ const pc = new Pinecone({ apiKey: 'YOUR_API_KEY' });
 const indexName = 'sample-index';
 
 // Ensure you have a serverless index
-await pc.createIndex({
+const indexModel = await pc.createIndex({
   name: indexName,
   dimension: 10,
   spec: {
@@ -35,7 +35,8 @@ await pc.createIndex({
   },
 });
 
-const index = pc.index({ name: indexName });
+// Get the host from the create response
+const index = pc.index({ host: indexModel.host });
 
 const storageURI = 's3://my-bucket/my-directory/';
 
@@ -58,7 +59,9 @@ You can check the status of an import operation using `describeImport`:
 import { Pinecone } from '@pinecone-database/pinecone';
 
 const pc = new Pinecone({ apiKey: 'YOUR_API_KEY' });
-const index = pc.index({ name: 'sample-index' });
+
+const indexModel = await pc.describeIndex('sample-index');
+const index = pc.index({ host: indexModel.host });
 
 const importStatus = await index.describeImport('import-id');
 console.log(importStatus);
@@ -81,7 +84,9 @@ You can list all import operations for an index:
 import { Pinecone } from '@pinecone-database/pinecone';
 
 const pc = new Pinecone({ apiKey: 'YOUR_API_KEY' });
-const index = pc.index({ name: 'sample-index' });
+
+const indexModel = await pc.describeIndex('sample-index');
+const index = pc.index({ host: indexModel.host });
 
 const imports = await index.listImports();
 console.log(imports);
@@ -124,7 +129,9 @@ You can cancel an ongoing import operation:
 import { Pinecone } from '@pinecone-database/pinecone';
 
 const pc = new Pinecone({ apiKey: 'YOUR_API_KEY' });
-const index = pc.index({ name: 'sample-index' });
+
+const indexModel = await pc.describeIndex('sample-index');
+const index = pc.index({ host: indexModel.host });
 
 await index.cancelImport('import-id');
 ```
@@ -139,7 +146,9 @@ When starting an import, you can specify how errors should be handled:
 import { Pinecone } from '@pinecone-database/pinecone';
 
 const pc = new Pinecone({ apiKey: 'YOUR_API_KEY' });
-const index = pc.index({ name: 'sample-index' });
+
+const indexModel = await pc.describeIndex('sample-index');
+const index = pc.index({ host: indexModel.host });
 
 // Continue on errors
 await index.startImport({
@@ -162,7 +171,9 @@ Here's a complete example of monitoring an import until completion:
 import { Pinecone } from '@pinecone-database/pinecone';
 
 const pc = new Pinecone({ apiKey: 'YOUR_API_KEY' });
-const index = pc.index({ name: 'sample-index' });
+
+const indexModel = await pc.describeIndex('sample-index');
+const index = pc.index({ host: indexModel.host });
 
 // Start the import
 const { id } = await index.startImport({

guides/data-operations/metadata-filtering.md

@@ -25,7 +25,9 @@ Pinecone supports the following filter operators:
 import { Pinecone } from '@pinecone-database/pinecone';
 
 const pc = new Pinecone({ apiKey: 'YOUR_API_KEY' });
-const index = pc.index({ name: 'my-index' });
+
+const indexModel = await pc.describeIndex('my-index');
+const index = pc.index({ host: indexModel.host });
 
 const results = await index.query({
   vector: [0.1, 0.2, 0.3, 0.4],
@@ -42,7 +44,9 @@ const results = await index.query({
 import { Pinecone } from '@pinecone-database/pinecone';
 
 const pc = new Pinecone({ apiKey: 'YOUR_API_KEY' });
-const index = pc.index({ name: 'my-index' });
+
+const indexModel = await pc.describeIndex('my-index');
+const index = pc.index({ host: indexModel.host });
 
 const results = await index.query({
   vector: [0.1, 0.2, 0.3, 0.4],
@@ -59,7 +63,9 @@ const results = await index.query({
 import { Pinecone } from '@pinecone-database/pinecone';
 
 const pc = new Pinecone({ apiKey: 'YOUR_API_KEY' });
-const index = pc.index({ name: 'my-index' });
+
+const indexModel = await pc.describeIndex('my-index');
+const index = pc.index({ host: indexModel.host });
 
 const results = await index.query({
   vector: [0.1, 0.2, 0.3, 0.4],
@@ -80,7 +86,9 @@ By default, multiple filter conditions are combined with AND:
 import { Pinecone } from '@pinecone-database/pinecone';
 
 const pc = new Pinecone({ apiKey: 'YOUR_API_KEY' });
-const index = pc.index({ name: 'my-index' });
+
+const indexModel = await pc.describeIndex('my-index');
+const index = pc.index({ host: indexModel.host });
 
 const results = await index.query({
   vector: [0.1, 0.2, 0.3, 0.4],
@@ -117,7 +125,9 @@ Use `$or` to match vectors that satisfy any of multiple conditions:
 import { Pinecone } from '@pinecone-database/pinecone';
 
 const pc = new Pinecone({ apiKey: 'YOUR_API_KEY' });
-const index = pc.index({ name: 'my-index' });
+
+const indexModel = await pc.describeIndex('my-index');
+const index = pc.index({ host: indexModel.host });
 
 const results = await index.query({
   vector: [0.1, 0.2, 0.3, 0.4],
@@ -142,7 +152,9 @@ You can nest `$and` and `$or` operators to create complex filters:
 import { Pinecone } from '@pinecone-database/pinecone';
 
 const pc = new Pinecone({ apiKey: 'YOUR_API_KEY' });
-const index = pc.index({ name: 'my-index' });
+
+const indexModel = await pc.describeIndex('my-index');
+const index = pc.index({ host: indexModel.host });
 
 const results = await index.query({
   vector: [0.1, 0.2, 0.3, 0.4],
@@ -167,7 +179,9 @@ Combine different field types and operators:
 import { Pinecone } from '@pinecone-database/pinecone';
 
 const pc = new Pinecone({ apiKey: 'YOUR_API_KEY' });
-const index = pc.index({ name: 'my-index' });
+
+const indexModel = await pc.describeIndex('my-index');
+const index = pc.index({ host: indexModel.host });
 
 const results = await index.query({
   vector: [0.1, 0.2, 0.3, 0.4],
@@ -190,6 +204,8 @@ import { Pinecone } from '@pinecone-database/pinecone';
 
 const pc = new Pinecone({ apiKey: 'YOUR_API_KEY' });
 
+const indexModel = await pc.describeIndex('my-index');
+
 type MovieMetadata = {
   genre: 'comedy' | 'horror' | 'drama' | 'action';
   year: number;

guides/data-operations/namespaces.md

@@ -2,29 +2,33 @@
 
 Namespaces allow you to partition vectors within an index. This is useful for organizing different types of data, implementing multitenancy, or isolating test data from production data.
 
-For more information, see [Use namespaces](https://docs.pinecone.io/guides/manage-data/manage-namespaces).
+For more information, see [Manage namespaces](https://docs.pinecone.io/guides/manage-data/manage-namespaces).
 
 ## Working with namespaces
 
-By default, all data operations take place inside the default namespace of `''`. If you are working with other non-default namespaces, you can specify the namespace when targeting an index:
+By default, all data operations take place inside the default namespace of `'__default__'`. If you are working with other non-default namespaces, you can specify the namespace when targeting an index:
 
 ```typescript
 import { Pinecone } from '@pinecone-database/pinecone';
 
 const pc = new Pinecone({ apiKey: 'YOUR_API_KEY' });
-const index = pc.index({ name: 'test-index', namespace: 'ns1' });
+
+const indexModel = await pc.describeIndex('test-index');
+const index = pc.index({ host: indexModel.host, namespace: 'ns1' });
 
 // Now all operations will use the 'ns1' namespace
 await index.fetch({ ids: ['1'] });
 ```
 
-Alternatively, you can specify the namespace for individual operations:
+Alternatively, you can specify the namespace for individual operations. Note that this will override any namespace set on the `Index` class:
 
 ```typescript
 import { Pinecone } from '@pinecone-database/pinecone';
 
 const pc = new Pinecone({ apiKey: 'YOUR_API_KEY' });
-const index = pc.index({ name: 'test-index' });
+
+const indexModel = await pc.describeIndex('test-index');
+const index = pc.index({ host: indexModel.host });
 
 // Query in a specific namespace
 await index.query({
@@ -42,7 +46,9 @@ The following example lists all namespaces in an index:
 import { Pinecone } from '@pinecone-database/pinecone';
 
 const pc = new Pinecone({ apiKey: 'YOUR_API_KEY' });
-const index = pc.index({ name: 'test-index' });
+
+const indexModel = await pc.describeIndex('test-index');
+const index = pc.index({ host: indexModel.host });
 
 const namespaces = await index.listNamespaces();
 console.log(namespaces);
@@ -59,7 +65,9 @@ The following example describes a specific namespace:
 import { Pinecone } from '@pinecone-database/pinecone';
 
 const pc = new Pinecone({ apiKey: 'YOUR_API_KEY' });
-const index = pc.index({ name: 'test-index' });
+
+const indexModel = await pc.describeIndex('test-index');
+const index = pc.index({ host: indexModel.host });
 
 const namespace = await index.describeNamespace('ns1');
 console.log(namespace);
@@ -77,7 +85,9 @@ The following example deletes a namespace and all its vectors:
 import { Pinecone } from '@pinecone-database/pinecone';
 
 const pc = new Pinecone({ apiKey: 'YOUR_API_KEY' });
-const index = pc.index({ name: 'test-index' });
+
+const indexModel = await pc.describeIndex('test-index');
+const index = pc.index({ host: indexModel.host });
 
 await index.deleteNamespace('ns1');
 ```
@@ -86,10 +96,9 @@ await index.deleteNamespace('ns1');
 
 ## Namespace best practices
 
-1. **Multitenancy**: Use namespaces to isolate data for different customers or tenants
+1. **Multitenancy**: Use namespaces to isolate data for different purposes
 2. **Environment separation**: Keep dev, staging, and production data separate
 3. **Data organization**: Group related vectors together (e.g., by document, user, or time period)
 4. **Prefix naming**: Consider using ID prefixes in combination with namespaces for hierarchical organization
-5. **Query performance**: Querying within a namespace is faster than querying across the entire index
 
 For more information on implementing multitenancy, see [Implement multitenancy](https://docs.pinecone.io/guides/index-data/implement-multitenancy).