@mastra/mssql 0.4.7 → 0.5.0

package/CHANGELOG.md

@@ -1,5 +1,65 @@
 # @mastra/mssql
 
+## 0.5.0
+
+### Minor Changes
+
+- Implemented AI tracing and observability features ([#9368](https://github.com/mastra-ai/mastra/pull/9368))
+  - Added createAISpan, updateAISpan, getAITrace, getAITracesPaginated
+  - Added batchCreateAISpans, batchUpdateAISpans, batchDeleteAITraces
+  - Automatic performance indexes for AI spans
+
+  Implemented workflow update methods
+  - Added updateWorkflowResults with row-level locking (UPDLOCK, HOLDLOCK)
+  - Added updateWorkflowState with row-level locking
+  - Concurrent update protection for parallel workflow execution
+
+  Added index management API
+  - Added createIndex, listIndexes, describeIndex, dropIndex methods
+  - Exposed index management methods directly on store instance
+  - Support for composite indexes, unique constraints, and filtered indexes
+
+  Documentation improvements
+  - Comprehensive README with complete API reference (58 methods)
+  - Detailed feature descriptions for all storage capabilities
+  - Index management examples and best practices
+  - Updated to reflect all atomic transaction usage
+
+### Patch Changes
+
+- Updated dependencies [[`f57a81e`](https://github.com/mastra-ai/mastra/commit/f57a81e6ce644e45bf1c9618778cc54c50a84ad4), [`2afd345`](https://github.com/mastra-ai/mastra/commit/2afd3450825b76e41f7973baddf13867ea042e40), [`fc79af3`](https://github.com/mastra-ai/mastra/commit/fc79af3915d1c456729cbd753673b0c0564340d8), [`eefc89e`](https://github.com/mastra-ai/mastra/commit/eefc89ee69f05bb71661473a807fc7dc03d56f17), [`60bd45d`](https://github.com/mastra-ai/mastra/commit/60bd45de021f0dfbe6583928f6da5169cb5585ba), [`a30093d`](https://github.com/mastra-ai/mastra/commit/a30093de98c1836dcd5dfddf09649010712b8c95), [`0fe7adb`](https://github.com/mastra-ai/mastra/commit/0fe7adb0f20f59a6bb41f235d01f8b7a880ea6e7), [`a42e496`](https://github.com/mastra-ai/mastra/commit/a42e49686a7486e2e9e9397fa98e5ff7a71dc1b0), [`3670db7`](https://github.com/mastra-ai/mastra/commit/3670db7e8e798f9d65fac5bfb732134a1f26ba7b), [`e40d4d0`](https://github.com/mastra-ai/mastra/commit/e40d4d0a0971b4505e7c9de73c656066c7565653), [`fc843ff`](https://github.com/mastra-ai/mastra/commit/fc843ff4d1d149317b6324553ce5ad7972062a78), [`ff16f9b`](https://github.com/mastra-ai/mastra/commit/ff16f9b9dbc701b26b6c4e9872f759f3880f9327), [`35e6cf7`](https://github.com/mastra-ai/mastra/commit/35e6cf722fef16ea0301eb9cf5a32fe9ccb12d22), [`30a2e36`](https://github.com/mastra-ai/mastra/commit/30a2e369485e0e59c4faa1d83c5635c2260b304c)]:
+  - @mastra/core@0.23.2
+
+## 0.5.0-alpha.0
+
+### Minor Changes
+
+- Implemented AI tracing and observability features ([#9368](https://github.com/mastra-ai/mastra/pull/9368))
+  - Added createAISpan, updateAISpan, getAITrace, getAITracesPaginated
+  - Added batchCreateAISpans, batchUpdateAISpans, batchDeleteAITraces
+  - Automatic performance indexes for AI spans
+
+  Implemented workflow update methods
+  - Added updateWorkflowResults with row-level locking (UPDLOCK, HOLDLOCK)
+  - Added updateWorkflowState with row-level locking
+  - Concurrent update protection for parallel workflow execution
+
+  Added index management API
+  - Added createIndex, listIndexes, describeIndex, dropIndex methods
+  - Exposed index management methods directly on store instance
+  - Support for composite indexes, unique constraints, and filtered indexes
+
+  Documentation improvements
+  - Comprehensive README with complete API reference (58 methods)
+  - Detailed feature descriptions for all storage capabilities
+  - Index management examples and best practices
+  - Updated to reflect all atomic transaction usage
+
+### Patch Changes
+
+- Updated dependencies [[`2afd345`](https://github.com/mastra-ai/mastra/commit/2afd3450825b76e41f7973baddf13867ea042e40), [`eefc89e`](https://github.com/mastra-ai/mastra/commit/eefc89ee69f05bb71661473a807fc7dc03d56f17), [`0fe7adb`](https://github.com/mastra-ai/mastra/commit/0fe7adb0f20f59a6bb41f235d01f8b7a880ea6e7), [`a42e496`](https://github.com/mastra-ai/mastra/commit/a42e49686a7486e2e9e9397fa98e5ff7a71dc1b0), [`3670db7`](https://github.com/mastra-ai/mastra/commit/3670db7e8e798f9d65fac5bfb732134a1f26ba7b), [`fc843ff`](https://github.com/mastra-ai/mastra/commit/fc843ff4d1d149317b6324553ce5ad7972062a78)]:
+  - @mastra/core@0.23.2-alpha.0
+
 ## 0.4.7
 
 ### Patch Changes

package/README.md

@@ -17,38 +17,84 @@ npm install @mastra/mssql
 
 ### Storage
 
+#### Basic Configuration
+
+MSSQLStore supports multiple connection methods:
+
+**1. Connection String (Recommended)**
+
 ```typescript
 import { MSSQLStore } from '@mastra/mssql';
 
+const store = new MSSQLStore({
+  connectionString:
+    'Server=localhost,1433;Database=mastra;User Id=sa;Password=yourPassword;Encrypt=true;TrustServerCertificate=true',
+});
+```
+
+**2. Server/Port/Database Configuration**
+
+```typescript
 const store = new MSSQLStore({
   server: 'localhost',
   port: 1433,
   database: 'mastra',
   user: 'sa',
   password: 'yourStrong(!)Password',
-  // options: { encrypt: true, trustServerCertificate: true }, // Optional
+  options: { encrypt: true, trustServerCertificate: true }, // Optional
 });
+```
 
-// Create a thread
-await store.saveThread({
-  id: 'thread-123',
-  resourceId: 'resource-456',
-  title: 'My Thread',
-  metadata: { key: 'value' },
+#### Advanced Options
+
+```typescript
+const store = new MSSQLStore({
+  connectionString:
+    'Server=localhost,1433;Database=mastra;User Id=sa;Password=yourPassword;Encrypt=true;TrustServerCertificate=true',
+  schemaName: 'custom_schema', // Use custom schema (default: dbo)
+  options: {
+    encrypt: true,
+    trustServerCertificate: true,
+    connectTimeout: 30000,
+    requestTimeout: 30000,
+    pool: {
+      max: 20,
+      min: 0,
+      idleTimeoutMillis: 30000,
+    },
+  },
 });
+```
 
-// Add messages to thread
-await store.saveMessages([
-  {
-    id: 'msg-789',
-    threadId: 'thread-123',
-    role: 'user',
-    type: 'text',
-    content: [{ type: 'text', text: 'Hello' }],
+#### Usage Example
+
+```typescript
+// Create a thread
+await store.saveThread({
+  thread: {
+    id: 'thread-123',
     resourceId: 'resource-456',
+    title: 'My Thread',
+    metadata: { key: 'value' },
     createdAt: new Date(),
+    updatedAt: new Date(),
   },
-]);
+});
+
+// Add messages to thread
+await store.saveMessages({
+  messages: [
+    {
+      id: 'msg-789',
+      threadId: 'thread-123',
+      role: 'user',
+      type: 'text',
+      content: [{ type: 'text', text: 'Hello' }],
+      resourceId: 'resource-456',
+      createdAt: new Date(),
+    },
+  ],
+});
 
 // Query threads and messages
 const savedThread = await store.getThreadById({ threadId: 'thread-123' });
@@ -57,38 +103,271 @@ const messages = await store.getMessages({ threadId: 'thread-123' });
 
 ## Configuration
 
-The MSSQL store can be initialized with either:
+### Connection Methods
+
+MSSQLStore supports multiple connection methods:
+
+1. **Connection String**
 
-- `connectionString`: Microsoft SQL Server connection string
-- Configuration object with server, port, database, user, and password
+   ```typescript
+   {
+     connectionString: 'Server=localhost,1433;Database=mastra;User Id=sa;Password=yourPassword;Encrypt=true;TrustServerCertificate=true';
+   }
+   ```
 
-Connection pool settings are managed by the [mssql](https://www.npmjs.com/package/mssql) package.
+2. **Server/Port/Database**
+   ```typescript
+   {
+     server: 'localhost',
+     port: 1433,
+     database: 'mastra',
+     user: 'sa',
+     password: 'password'
+   }
+   ```
+
+### Optional Configuration
+
+- `schemaName`: Custom SQL Server schema (default: `dbo`)
+- `options.encrypt`: Enable encryption (default: `true`)
+- `options.trustServerCertificate`: Trust self-signed certificates (default: `true`)
+- `options.connectTimeout`: Connection timeout in milliseconds (default: `15000`)
+- `options.requestTimeout`: Request timeout in milliseconds (default: `15000`)
+- `options.pool.max`: Maximum pool connections (default: `10`)
+- `options.pool.min`: Minimum pool connections (default: `0`)
+- `options.pool.idleTimeoutMillis`: Idle connection timeout (default: `30000`)
+
+### Default Connection Pool Settings
+
+- Maximum connections: 10
+- Minimum connections: 0
+- Idle timeout: 30 seconds
+- Connection timeout: 15 seconds
+- Request timeout: 15 seconds
 
 ## Features
 
-- Thread and message storage with JSON support
-- Atomic transactions for data consistency
-- Efficient batch operations
-- Rich metadata support
-- Timestamp tracking
-- Cascading deletes (emulated)
+### Storage Features
+
+- **Thread and Message Management**
+  - Thread and message storage with JSON support
+  - Message format versioning (v1 and v2)
+  - Pagination support for threads and messages
+  - Atomic transactions for batch operations (save, update, delete)
+  - Automatic thread timestamp updates
+  - Cascading deletes
+- **Resources**
+  - Resource storage with working memory
+  - Rich metadata support
+  - Update working memory and metadata independently
+
+- **AI Tracing & Observability**
+  - Trace AI agent execution with spans
+  - Query traces with pagination and filtering
+  - Batch operations for high-volume tracing
+  - Parent-child span relationships
+  - Span metadata and timing information
+
+- **Workflow Management**
+  - Persist and restore workflow execution state
+  - Track workflow run history
+  - Step-by-step result tracking with row-level locking
+  - Workflow status management with row-level locking
+  - Query workflow runs by date range or resource
+  - Concurrent update protection for parallel workflow execution
+
+- **Scoring & Evaluation**
+  - Store evaluation scores and metrics
+  - Query scores by scorer, run, entity, or span
+  - Support for multiple scoring sources
+  - Pagination support for large score datasets
+
+- **Performance & Scalability**
+  - Connection pooling with configurable limits
+  - Atomic transactions for all batch operations
+  - Efficient batch insert/update/delete with transaction safety
+  - Row-level locking for concurrent updates
+  - Automatic performance indexes
+  - Index management (create, list, describe, drop)
+  - Timestamp tracking with high precision
+- **Data Management**
+  - Custom schema support
+  - Table operations (create, alter, clear, drop)
+  - Low-level insert and load operations
+  - JSON data type support
 
 ## Storage Methods
 
+### Initialization & Connection
+
+- `init()`: Initialize the store and create tables
+- `close()`: Close database connection pool
+
+### Threads
+
 - `saveThread({ thread })`: Create or update a thread
 - `getThreadById({ threadId })`: Get a thread by ID
+- `updateThread({ id, title, metadata })`: Update thread title and metadata
 - `deleteThread({ threadId })`: Delete a thread and its messages
-- `saveMessages({ messages })`: Save multiple messages in a transaction
-- `getMessages({ threadId })`: Get all messages for a thread
-- `updateMessages({ messages })`: Update messages by ID
-- `getThreadsByResourceIdPaginated({ resourceId, page, perPage })`: Paginated thread listing
-- `getMessagesPaginated({ threadId, selectBy })`: Paginated message listing
-- `clearTable({ tableName })`: Remove all rows from a table (with cascade)
-- `createTable({ tableName, schema })`: Create a table if it does not exist
-- `alterTable({ tableName, schema, ifNotExists })`: Add columns if they do not exist
-- `saveResource({ resource })`: Save a resource
+- `getThreadsByResourceIdPaginated({ resourceId, page, perPage, orderBy?, sortDirection? })`: Get paginated threads for a resource
+
+### Messages
+
+- `saveMessages({ messages, format? })`: Save multiple messages with atomic transaction (supports v1 and v2 formats)
+- `getMessages({ threadId, format? })`: Get all messages for a thread
+- `getMessagesById({ messageIds, format? })`: Get messages by their IDs
+- `getMessagesPaginated({ threadId, format?, page?, perPage? })`: Get paginated messages for a thread
+- `updateMessages({ messages })`: Update existing messages with atomic transaction
+- `deleteMessages(messageIds)`: Delete specific messages with atomic transaction
+
+### Resources
+
+- `saveResource({ resource })`: Save a resource with working memory
 - `getResourceById({ resourceId })`: Get a resource by ID
-- `updateResource({ resourceId, ... })`: Update a resource
+- `updateResource({ resourceId, workingMemory?, metadata? })`: Update resource working memory and metadata
+
+### AI Tracing & Observability
+
+- `createAISpan(span)`: Create an AI trace span
+- `updateAISpan({ spanId, traceId, updates })`: Update an existing span
+- `getAITrace(traceId)`: Get complete trace with all spans
+- `getAITracesPaginated({ filters?, pagination? })`: Query traces with pagination and filters
+- `batchCreateAISpans({ records })`: Batch create multiple spans
+- `batchUpdateAISpans({ records })`: Batch update multiple spans
+- `batchDeleteAITraces({ traceIds })`: Batch delete traces
+
+### Index Management
+
+- `createIndex({ name, table, columns, unique?, where? })`: Create a new index
+- `listIndexes(tableName?)`: List all indexes or indexes for a specific table
+- `describeIndex(indexName)`: Get detailed index statistics and information
+- `dropIndex(indexName)`: Drop an existing index
+
+### Workflows
+
+- `persistWorkflowSnapshot({ workflowName, runId, resourceId?, snapshot })`: Save workflow execution state
+- `loadWorkflowSnapshot({ workflowName, runId })`: Load workflow execution state
+- `updateWorkflowResults({ workflowName, runId, stepId, result, runtimeContext })`: Update step results (transaction + row locking)
+- `updateWorkflowState({ workflowName, runId, opts })`: Update workflow run status (transaction + row locking)
+- `getWorkflowRuns({ workflowName?, fromDate?, toDate?, limit?, offset?, resourceId? })`: Query workflow runs
+- `getWorkflowRunById({ runId, workflowName? })`: Get specific workflow run
+
+### Scores & Evaluation
+
+- `saveScore(score)`: Save evaluation score
+- `getScoreById({ id })`: Get score by ID
+- `getScoresByScorerId({ scorerId, pagination, entityId?, entityType?, source? })`: Get scores by scorer
+- `getScoresByRunId({ runId, pagination })`: Get scores for a run
+- `getScoresByEntityId({ entityId, entityType, pagination })`: Get scores for an entity
+- `getScoresBySpan({ traceId, spanId, pagination })`: Get scores for a trace span
+
+### Traces (Legacy)
+
+- `getTracesPaginated({ filters?, pagination? })`: Get paginated legacy traces
+- `batchTraceInsert({ records })`: Batch insert legacy trace records
+
+### Evals (Legacy)
+
+- `getEvals({ agentName?, type?, page?, perPage? })`: Get paginated evaluations
+
+### Low-level Operations
+
+- `createTable({ tableName, schema })`: Create a new table
+- `alterTable({ tableName, schema, ifNotExists })`: Add columns to existing table
+- `clearTable({ tableName })`: Remove all rows from a table
+- `dropTable({ tableName })`: Drop a table
+- `insert({ tableName, record })`: Insert a single record
+- `batchInsert({ tableName, records })`: Batch insert multiple records
+- `load<R>({ tableName, keys })`: Load a record by key(s)
+
+## Index Management
+
+The MSSQL store provides comprehensive index management capabilities to optimize query performance.
+
+### Automatic Performance Indexes
+
+MSSQL storage automatically creates composite indexes during initialization for common query patterns. These indexes significantly improve performance for filtered queries with sorting.
+
+### Creating Custom Indexes
+
+```typescript
+// Basic index for common queries
+await store.createIndex({
+  name: 'idx_threads_resource',
+  table: 'mastra_threads',
+  columns: ['resourceId'],
+});
+
+// Composite index with sort order for filtering + sorting
+await store.createIndex({
+  name: 'idx_messages_composite',
+  table: 'mastra_messages',
+  columns: ['thread_id', 'seq_id DESC'],
+});
+
+// Unique index for constraints
+await store.createIndex({
+  name: 'idx_unique_constraint',
+  table: 'mastra_resources',
+  columns: ['id'],
+  unique: true,
+});
+
+// Filtered index (partial indexing)
+await store.createIndex({
+  name: 'idx_active_threads',
+  table: 'mastra_threads',
+  columns: ['resourceId'],
+  where: "status = 'active'",
+});
+```
+
+### Managing Indexes
+
+```typescript
+// List all indexes
+const allIndexes = await store.listIndexes();
+
+// List indexes for specific table
+const threadIndexes = await store.listIndexes('mastra_threads');
+
+// Get detailed statistics for an index
+const stats = await store.describeIndex('idx_threads_resource');
+console.log(stats);
+// {
+//   name: 'idx_threads_resource',
+//   table: 'mastra_threads',
+//   columns: ['resourceId', 'seq_id'],
+//   unique: false,
+//   size: '128 KB',
+//   method: 'nonclustered',
+//   scans: 1542,           // Number of index seeks
+//   tuples_read: 45230,    // Tuples read via index
+//   tuples_fetched: 12050  // Tuples fetched via index
+// }
+
+// Drop an index
+await store.dropIndex('idx_threads_resource');
+```
+
+### Monitoring Index Performance
+
+```typescript
+// Check index usage statistics
+const stats = await store.describeIndex('idx_threads_resource');
+
+// Identify unused indexes
+if (stats.scans === 0) {
+  console.log(`Index ${stats.name} is unused - consider removing`);
+  await store.dropIndex(stats.name);
+}
+
+// Monitor index efficiency
+const efficiency = stats.tuples_fetched / stats.tuples_read;
+if (efficiency < 0.5) {
+  console.log(`Index ${stats.name} has low efficiency: ${efficiency}`);
+}
+```
 
 ## Related Links