Knowledge for drafts and clean schema clean up

.cursor/rules/cleaner.mdc → .cursor/rules/features/cleaner.mdc

@@ -5,6 +5,8 @@ alwaysApply: false
 ---
 ## Inbox Cleaner
 
+This file explains the Inbox Cleaner feature and how it's implemented.
+
 The inbox cleaner helps users do a deep clean of their inbox.
 It helps them get from 10,000 items in their inbox to only a few.
 It works by archiving/marking read low priority emails.

.cursor/rules/features/knowledge.mdc

@@ -0,0 +1,92 @@
+---
+description: 
+globs: 
+alwaysApply: false
+---
+# Knowledge Base
+
+This file explains the Knowledge Base feature and how it's implemented.
+
+The knowledge base helps users store and manage information that can be used to help draft responses to emails. It acts as a personal database of information that can be referenced when composing replies.
+
+## Overview
+
+Users can create, edit, and delete knowledge base entries. Each entry consists of:
+
+- A title for quick reference
+- Content that contains the actual information
+- Metadata like creation and update timestamps
+
+## Database Schema
+
+The `Knowledge` model in Prisma:
+
+```prisma
+model Knowledge {
+  id        String   @id @default(cuid())
+  createdAt DateTime @default(now())
+  updatedAt DateTime @updatedAt
+  title     String
+  content   String
+
+  userId String
+  user   User   @relation(fields: [userId], references: [id], onDelete: Cascade)
+}
+```
+
+Each knowledge entry belongs to a specific user and is automatically deleted if the user is deleted (cascade).
+
+## Main Files and Directories
+
+The knowledge base functionality is implemented in:
+
+- `apps/web/app/(app)/automation/knowledge/KnowledgeBase.tsx` - Main UI component
+- `apps/web/app/(app)/automation/knowledge/KnowledgeForm.tsx` - Form for creating/editing entries
+- `apps/web/utils/actions/knowledge.ts` - Server actions for CRUD operations
+- `apps/web/utils/actions/knowledge.validation.ts` - Zod validation schemas
+- `apps/web/app/api/knowledge/route.ts` - API route for fetching entries
+
+### AI Integration Files
+
+- `apps/web/utils/ai/knowledge/extract.ts` - Extract relevant knowledge from knowledge base entries
+- `apps/web/utils/ai/knowledge/extract-from-email-history.ts` - Extract context from previous emails
+- `apps/web/utils/ai/reply/draft-with-knowledge.ts` - Generate email drafts using extracted knowledge
+- `apps/web/utils/reply-tracker/generate-reply.ts` - Coordinates the extraction and drafting process
+- `apps/web/utils/llms/model-selector.ts` - Economy LLM selection for high-volume tasks
+
+## Features
+
+- **Create**: Users can add new knowledge entries with a title and content
+- **Read**: Entries are displayed in a table with title and last updated date
+- **Update**: Users can edit existing entries
+- **Delete**: Entries can be deleted with a confirmation dialog
+
+## Usage in Email Responses
+
+The knowledge base entries are used to help draft responses to emails. When composing a reply, the system can reference these entries to include relevant information, ensuring consistent and accurate responses.
+
+When drafting responses, we use two LLMs:
+
+1. A cheaper LLM that can process a lot of data (e.g. Google Gemini 2 Flash)
+2. A more expensive LLM to draft the response (e.g. Anthropic Sonnet 3.7)
+
+The cheaper LLM is an agent that extracts the key information needed for the drafter LLM.
+For example, the knowledge base may include 100 pages of content, and the LLM extracts half a page of knowledge to pass to the more expensive drafter LLM.
+
+## Dual LLM Architecture
+
+The dual LLM approach is implemented as follows:
+
+1. **Knowledge Extraction (Economy LLM)**:
+
+   - Uses a more cost-efficient model like Gemini Flash for processing large volumes of knowledge base content
+   - Analyzes all knowledge entries and extracts only relevant information based on the email content
+   - Configured via environment variables (`ECONOMY_LLM_PROVIDER` and `ECONOMY_LLM_MODEL`)
+   - If no specific economy model is configured, defaults to Gemini Flash when Google API key is available
+
+2. **Email Draft Generation (Core LLM)**:
+   - Uses the default model (e.g., Anthropic Claude 3.7 Sonnet) for high-quality content generation
+   - Receives the extracted relevant knowledge from the economy LLM
+   - Generates the final email draft based on the provided context
+
+This architecture optimizes for both cost efficiency (using cheaper models for high-volume tasks) and quality (using premium models for user-facing content).

.cursor/rules/features/reply-tracker.mdc

@@ -0,0 +1,25 @@
+---
+description: 
+globs: 
+alwaysApply: false
+---
+# Reply Tracker
+
+Reply Tracker (also known as Reply Zero) lets the user which emails need a reply for them and those they're awaiting a reply on.
+It updates the labels for the thread automatically in Gmail.
+
+The database models and fields that are used for this feature:
+
+- ThreadTracker
+- User.outboundReplyTracking
+- ActionType.TRACK_THREAD
+
+The system uses rules. The AI can choose which rule to use each time an email comes in. Rules are for incoming emails only. Not ongoing.
+When enabling the reply tracker, we create a rule for the user that has the following actions associated with it:
+- LABEL: "To Reply"
+- TRACK_THREAD
+- DRAFT_EMAIL (optional)
+
+We'll draft a reply for the user automatically if DRAFT_EMAIL is set. The draft will be generated using the email history for this sender as well as the Knowledge base. See `.cursor/rules/features/knowledge.mdc` for more on the Knowledge Base feature.
+
+Enabling `User.outboundReplyTracking` means that when a user sends an email, we'll run an LLM over the email and check if it needs a reply. If it does we'll mark it as Awaiting Reply.

.cursor/rules/installing-packages.mdc

@@ -0,0 +1,12 @@
+---
+description: How to install packages
+globs: 
+alwaysApply: false
+---
+- Use `pnpm`.
+- Don't install in root. Install in `apps/web`:
+
+```sh
+cd apps/web
+pnpm add ...
+```

.cursor/rules/llm-test.mdc

@@ -13,9 +13,8 @@ Tests for LLM-related functionality should follow these guidelines to ensure con
 
    ```
    apps/web/__tests__/
-   ├── ai/
-   │   └── your-feature.test.ts
-   │   └── another-feature.test.ts
+   │ └── your-feature.test.ts
+   │ └── another-feature.test.ts
    └── ...
    ```
 
@@ -25,7 +24,8 @@ Tests for LLM-related functionality should follow these guidelines to ensure con
    import { describe, expect, test, vi, beforeEach } from "vitest";
    import { yourFunction } from "@/utils/ai/your-feature";
 
-   // Required for Next.js compatibility
+   // Run with: pnpm test-ai TEST
+
    vi.mock("server-only", () => ({}));
 
    // Skip tests unless explicitly running AI tests

apps/web/__tests__/ai-choose-args.test.ts

@@ -1,7 +1,7 @@
 import { describe, expect, test, vi } from "vitest";
-import { getActionItemsWithAiArgs } from "@/utils/ai/choose-rule/ai-choose-args";
 import { type Action, ActionType, LogicalOperator } from "@prisma/client";
-import type { RuleWithActions } from "@/utils/types";
+import type { ParsedMessage, RuleWithActions } from "@/utils/types";
+import { getActionItemsWithAiArgs } from "@/utils/ai/choose-rule/choose-args";
 
 // pnpm test-ai ai-choose-args
 
@@ -15,9 +15,13 @@ describe.skipIf(!isAiTest)("getActionItemsWithAiArgs", () => {
     const rule = getRule("Test rule", actions);
 
     const result = await getActionItemsWithAiArgs({
-      email: getEmail(),
+      message: getParsedMessage({
+        subject: "Test subject",
+        content: "Test content",
+      }),
       user: getUser(),
       selectedRule: rule,
+      gmail: {} as any,
     });
 
     expect(result).toEqual(actions);
@@ -33,12 +37,13 @@ describe.skipIf(!isAiTest)("getActionItemsWithAiArgs", () => {
     const rule = getRule("Choose this rule for meeting requests", actions);
 
     const result = await getActionItemsWithAiArgs({
-      email: getEmail({
+      message: getParsedMessage({
         subject: "Quick question",
         content: "When is the meeting tomorrow?",
       }),
       user: getUser(),
       selectedRule: rule,
+      gmail: {} as any,
     });
 
     expect(result).toHaveLength(1);
@@ -59,12 +64,13 @@ describe.skipIf(!isAiTest)("getActionItemsWithAiArgs", () => {
     );
 
     const result = await getActionItemsWithAiArgs({
-      email: getEmail({
+      message: getParsedMessage({
         subject: "Quick question",
         content: "How much are pears?",
       }),
       user: getUser(),
       selectedRule: rule,
+      gmail: {} as any,
     });
 
     expect(result).toHaveLength(1);
@@ -85,12 +91,13 @@ describe.skipIf(!isAiTest)("getActionItemsWithAiArgs", () => {
     const rule = getRule("Test rule", actions);
 
     const result = await getActionItemsWithAiArgs({
-      email: getEmail({
+      message: getParsedMessage({
         subject: "Project status",
         content: "Can you update me on the project status?",
       }),
       user: getUser(),
       selectedRule: rule,
+      gmail: {} as any,
     });
 
     expect(result).toHaveLength(2);
@@ -120,13 +127,14 @@ Matt`,
     );
 
     const result = await getActionItemsWithAiArgs({
-      email: getEmail({
+      message: getParsedMessage({
         from: "[email protected]",
         subject: "fruits",
         content: "how much do apples cost?",
       }),
       user: getUser(),
       selectedRule: rule,
+      gmail: {} as any,
     });
 
     expect(result).toHaveLength(2);
@@ -190,24 +198,40 @@ function getRule(
     categoryFilterType: null,
     conditionalOperator: LogicalOperator.AND,
     type: null,
-    trackReplies: null,
   };
 }
 
-function getEmail({
+function getParsedMessage({
   from = "[email protected]",
   subject = "subject",
   content = "content",
-}: { from?: string; subject?: string; content?: string } = {}) {
+}): ParsedMessage {
   return {
-    from,
-    subject,
-    content,
+    id: "id",
+    threadId: "thread-id",
+    snippet: "",
+    attachments: [],
+    historyId: "history-id",
+    sizeEstimate: 100,
+    internalDate: new Date().toISOString(),
+    inline: [],
+    textPlain: content,
+    // ...message,
+    headers: {
+      from,
+      to: "[email protected]",
+      subject,
+      date: new Date().toISOString(),
+      references: "",
+      "message-id": "message-id",
+      // ...message.headers,
+    },
   };
 }
 
 function getUser() {
   return {
+    id: "userId",
     aiModel: null,
     aiProvider: null,
     email: "[email protected]",