chore: sync lockfile and payload-ai extensions for release v1.9.10

packages/payload-ai/src/tools/mcpAdapter.ts

@@ -0,0 +1,65 @@
+import { Client } from '@modelcontextprotocol/sdk/client/index.js'
+import { SSEClientTransport } from '@modelcontextprotocol/sdk/client/sse.js'
+import { StdioClientTransport } from '@modelcontextprotocol/sdk/client/stdio.js'
+import { tool } from 'ai'
+import { z } from 'zod'
+
+/**
+ * Connects to an external MCP Server and maps its tools to Vercel AI SDK Tools.
+ */
+export async function createMcpTools(mcpConfig: { name: string, url?: string, command?: string, args?: string[] }) {
+    let transport
+
+    // Support both HTTP/SSE and STDIO transports
+    if (mcpConfig.url) {
+        transport = new SSEClientTransport(new URL(mcpConfig.url))
+    } else if (mcpConfig.command) {
+        transport = new StdioClientTransport({
+            command: mcpConfig.command,
+            args: mcpConfig.args || [],
+        })
+    } else {
+        throw new Error('Invalid MCP config: Must provide either URL or Command.')
+    }
+
+    const client = new Client(
+        { name: `payload-ai-client-${mcpConfig.name}`, version: '1.0.0' },
+        { capabilities: {} }
+    )
+
+    await client.connect(transport)
+
+    // Fetch available tools from the external MCP server
+    const toolListResult = await client.listTools()
+    const externalTools = toolListResult.tools || []
+
+    const aiSdkTools: Record<string, any> = {}
+
+    // Map each external tool to a Vercel AI SDK Tool
+    for (const extTool of externalTools) {
+        // Basic conversion of JSON Schema to Zod for the AI SDK
+        // Note: For a production ready adapter, you might need a more robust jsonSchemaToZod converter
+        // or use AI SDK's new experimental generateSchema feature if available.
+        // Here we use a generic `z.any()` as a fallback since AI SDK requires a Zod schema.
+        const toolSchema = extTool.inputSchema as Record<string, any>
+
+        // We create a simplified parameter parser. 
+        // An ideal approach uses `jsonSchemaToZod` library or native AI SDK JSON schema support
+        // (introduced recently in `ai` package). 
+
+        aiSdkTools[`${mcpConfig.name}_${extTool.name}`] = tool({
+            description: `[From ${mcpConfig.name}] ${extTool.description || extTool.name}`,
+            parameters: z.any().describe('JSON matching the original MCP input_schema'), // Simplify for prototype
+            // @ts-ignore - AI strict mode overload bug with implicit zod inferences
+            execute: async (args: any) => {
+                const result = await client.callTool({
+                    name: extTool.name,
+                    arguments: args
+                })
+                return result
+            }
+        })
+    }
+
+    return { tools: aiSdkTools, client }
+}

packages/payload-ai/src/tools/memoryDb.ts

@@ -0,0 +1,115 @@
+import { tool } from 'ai'
+import { z } from 'zod'
+import { QdrantClient } from '@qdrant/js-client-rest'
+
+// Qdrant initialization
+// This requires the user to have Qdrant running and QDRANT_URL/QDRANT_API_KEY environment variables set
+const qdrantClient = new QdrantClient({
+    url: process.env.QDRANT_URL || 'http://localhost:6333',
+    apiKey: process.env.QDRANT_API_KEY,
+})
+
+const MEMORY_COLLECTION = 'mintel_ai_memory'
+
+// Ensure collection exists on load
+async function initQdrant() {
+    try {
+        const res = await qdrantClient.getCollections()
+        const exists = res.collections.find((c: any) => c.name === MEMORY_COLLECTION)
+        if (!exists) {
+            await qdrantClient.createCollection(MEMORY_COLLECTION, {
+                vectors: {
+                    size: 1536, // typical embedding size, adjust based on the embedding model used
+                    distance: 'Cosine',
+                },
+            })
+            console.log(`Qdrant collection '${MEMORY_COLLECTION}' created.`)
+        }
+    } catch (error) {
+        console.error('Failed to initialize Qdrant memory collection:', error)
+    }
+}
+
+// Call init, but don't block
+initQdrant()
+
+/**
+ * Returns memory tools for the AI SDK.
+ * Note: A real implementation would require an embedding step before inserting into Qdrant.
+ * For this implementation, we use a placeholder or assume the embeddings are handled
+ * by a utility function, or we use Qdrant's FastEmbed (if running their specialized container).
+ */
+export const generateMemoryTools = (userId: string | number) => {
+    return {
+        save_memory: tool({
+            description: 'Save an important preference, fact, or instruction about the user to long-term memory. Only use this when explicitly asked or when it is clearly a long-term preference.',
+            parameters: z.object({
+                fact: z.string().describe('The fact or instruction to remember.'),
+                category: z.string().optional().describe('An optional category like "preference", "rule", or "project_detail".'),
+            }),
+            // @ts-ignore - AI SDK strict mode bug
+            execute: async ({ fact, category }: { fact: string; category?: string }) => {
+                // In a real scenario, you MUST generate embeddings for the 'fact' string here
+                // using OpenAI or another embedding provider before inserting into Qdrant.
+                // const embedding = await generateEmbedding(fact)
+
+                try {
+                    // Mock embedding payload for demonstration
+                    const mockEmbedding = new Array(1536).fill(0).map(() => Math.random())
+
+                    await qdrantClient.upsert(MEMORY_COLLECTION, {
+                        wait: true,
+                        points: [
+                            {
+                                id: crypto.randomUUID(),
+                                vector: mockEmbedding,
+                                payload: {
+                                    userId: String(userId), // Partition memory by user
+                                    fact,
+                                    category,
+                                    createdAt: new Date().toISOString(),
+                                },
+                            },
+                        ],
+                    })
+                    return { success: true, message: `Successfully remembered: "${fact}"` }
+                } catch (error) {
+                    console.error("Qdrant save error:", error)
+                    return { success: false, error: 'Failed to save to memory database.' }
+                }
+            },
+        }),
+
+        search_memory: tool({
+            description: 'Search the user\'s long-term memory for past factual context, preferences, or rules.',
+            parameters: z.object({
+                query: z.string().describe('The search string to find in memory.'),
+            }),
+            // @ts-ignore - AI SDK strict mode bug
+            execute: async ({ query }: { query: string }) => {
+                // Generate embedding for query
+                const mockQueryEmbedding = new Array(1536).fill(0).map(() => Math.random())
+
+                try {
+                    const results = await qdrantClient.search(MEMORY_COLLECTION, {
+                        vector: mockQueryEmbedding,
+                        limit: 5,
+                        filter: {
+                            must: [
+                                {
+                                    key: 'userId',
+                                    match: { value: String(userId) }
+                                }
+                            ]
+                        }
+                    })
+
+                    return results.map((r: any) => r.payload?.fact || '')
+                } catch (error) {
+                    console.error("Qdrant search error:", error)
+                    return []
+                }
+            }
+        })
+    }
+}

packages/payload-ai/src/tools/payloadLocal.ts

@@ -0,0 +1,107 @@
+import { tool } from 'ai'
+import { z } from 'zod'
+import type { Payload, PayloadRequest, User } from 'payload'
+
+export const generatePayloadLocalTools = (
+    payload: Payload,
+    req: PayloadRequest,
+    allowedCollections: string[]
+) => {
+    const tools: Record<string, any> = {}
+
+    for (const collectionSlug of allowedCollections) {
+        const slugKey = collectionSlug.replace(/-/g, '_')
+
+        // 1. Read (Find) Tool
+        tools[`read_${slugKey}`] = tool({
+            description: `Read/Find documents from the Payload CMS collection: ${collectionSlug}`,
+            parameters: z.object({
+                limit: z.number().optional().describe('Number of documents to return, max 100.'),
+                page: z.number().optional().describe('Page number for pagination.'),
+                // Simple string-based query for demo purposes. For a robust implementation,
+                // we'd map this to Payload's where query logic using a structured Zod schema.
+                query: z.string().optional().describe('Optional text to search within the collection.'),
+            }),
+            // @ts-ignore - AI SDK strict mode type inference bug
+            execute: async ({ limit = 10, page = 1, query }: { limit?: number; page?: number; query?: string }) => {
+                const where = query ? { id: { equals: query } } : undefined // Placeholder logic
+
+                return await payload.find({
+                    collection: collectionSlug as any,
+                    limit: Math.min(limit, 100),
+                    page,
+                    where,
+                    req, // Crucial for passing the user context and respecting access control!
+                })
+            },
+        })
+
+        // 2. Read by ID Tool
+        tools[`read_${slugKey}_by_id`] = tool({
+            description: `Get a specific document by its ID from the ${collectionSlug} collection.`,
+            parameters: z.object({
+                id: z.union([z.string(), z.number()]).describe('The ID of the document.'),
+            }),
+            // @ts-ignore - AI SDK strict mode type inference bug
+            execute: async ({ id }: { id: string | number }) => {
+                return await payload.findByID({
+                    collection: collectionSlug as any,
+                    id,
+                    req, // Enforce access control
+                })
+            },
+        })
+
+        // 3. Create Tool
+        tools[`create_${slugKey}`] = tool({
+            description: `Create a new document in the ${collectionSlug} collection.`,
+            parameters: z.object({
+                data: z.record(z.any()).describe('A JSON object containing the data to insert.'),
+            }),
+            // @ts-ignore - AI SDK strict mode type inference bug
+            execute: async ({ data }: { data: Record<string, any> }) => {
+                return await payload.create({
+                    collection: collectionSlug as any,
+                    data,
+                    req, // Enforce access control
+                })
+            },
+        })
+
+        // 4. Update Tool
+        tools[`update_${slugKey}`] = tool({
+            description: `Update an existing document in the ${collectionSlug} collection.`,
+            parameters: z.object({
+                id: z.union([z.string(), z.number()]).describe('The ID of the document to update.'),
+                data: z.record(z.any()).describe('A JSON object containing the fields to update.'),
+            }),
+            // @ts-ignore - AI SDK strict mode type inference bug
+            execute: async ({ id, data }: { id: string | number; data: Record<string, any> }) => {
+                return await payload.update({
+                    collection: collectionSlug as any,
+                    id,
+                    data,
+                    req, // Enforce access control
+                })
+            },
+        })
+
+        // 5. Delete Tool
+        tools[`delete_${slugKey}`] = tool({
+            description: `Delete a document from the ${collectionSlug} collection by ID.`,
+            parameters: z.object({
+                id: z.union([z.string(), z.number()]).describe('The ID of the document to delete.'),
+            }),
+            // @ts-ignore - AI SDK strict mode type inference bug
+            execute: async ({ id }: { id: string | number }) => {
+                return await payload.delete({
+                    collection: collectionSlug as any,
+                    id,
+                    req, // Enforce access control
+                })
+            },
+        })
+    }
+
+    return tools
+}