Merge branch 'main' of https://github.com/Dhravya/supermemory into new-uinew-ui

4 files changed, 114 insertions, 97 deletions

diff --git a/apps/cf-ai-backend/src/routes/add.ts b/apps/cf-ai-backend/src/routes/add.ts
index 9b05e9f0..b9a6da8f 100644
--- a/apps/cf-ai-backend/src/routes/add.ts
+++ b/apps/cf-ai-backend/src/routes/add.ts
@@ -1,36 +1,38 @@
-import { Request } from "@cloudflare/workers-types";
-import { type CloudflareVectorizeStore } from "@langchain/cloudflare";
+import { Request } from '@cloudflare/workers-types';
+import { type CloudflareVectorizeStore } from '@langchain/cloudflare';
 
 export async function POST(request: Request, store: CloudflareVectorizeStore) {
-	const body = await request.json() as {
-		pageContent: string,
-		title?: string,
-		description?: string,
-		space?: string,
-		url: string,
-		user: string
+	const body = (await request.json()) as {
+		pageContent: string;
+		title?: string;
+		description?: string;
+		space?: string;
+		url: string;
+		user: string;
 	};
 
 	if (!body.pageContent || !body.url) {
-		return new Response(JSON.stringify({ message: "Invalid Page Content" }), { status: 400 });
+		return new Response(JSON.stringify({ message: 'Invalid Page Content' }), { status: 400 });
 	}
-	const newPageContent = `Title: ${body.title}\nDescription: ${body.description}\nURL: ${body.url}\nContent: ${body.pageContent}`
+	const newPageContent = `Title: ${body.title}\nDescription: ${body.description}\nURL: ${body.url}\nContent: ${body.pageContent}`;
 
-
-	await store.addDocuments([
-		{
-			pageContent: newPageContent,
-			metadata: {
-				title: body.title ?? "",
-				description: body.description ?? "",
-				space: body.space ?? "",
-				url: body.url,
-				user: body.user,
+	await store.addDocuments(
+		[
+			{
+				pageContent: newPageContent,
+				metadata: {
+					title: body.title ?? '',
+					description: body.description ?? '',
+					space: body.space ?? '',
+					url: body.url,
+					user: body.user,
+				},
 			},
+		],
+		{
+			ids: [`${body.url}-${body.user}`],
 		},
-	], {
-		ids: [`${body.url}`]
-	})
+	);
 
-	return new Response(JSON.stringify({ message: "Document Added" }), { status: 200 });
+	return new Response(JSON.stringify({ message: 'Document Added' }), { status: 200 });
 }
diff --git a/apps/cf-ai-backend/src/routes/ask.ts b/apps/cf-ai-backend/src/routes/ask.ts
index 1c48dde8..267c1513 100644
--- a/apps/cf-ai-backend/src/routes/ask.ts
+++ b/apps/cf-ai-backend/src/routes/ask.ts
@@ -1,18 +1,18 @@
-import { GenerativeModel } from "@google/generative-ai";
-import { OpenAIEmbeddings } from "../OpenAIEmbedder";
-import { CloudflareVectorizeStore } from "@langchain/cloudflare";
-import { Request } from "@cloudflare/workers-types";
+import { GenerativeModel } from '@google/generative-ai';
+import { OpenAIEmbeddings } from '../OpenAIEmbedder';
+import { CloudflareVectorizeStore } from '@langchain/cloudflare';
+import { Request } from '@cloudflare/workers-types';
 
 export async function POST(request: Request, _: CloudflareVectorizeStore, embeddings: OpenAIEmbeddings, model: GenerativeModel, env?: Env) {
-	const body = await request.json() as {
-		query: string
+	const body = (await request.json()) as {
+		query: string;
 	};
 
 	if (!body.query) {
-		return new Response(JSON.stringify({ message: "Invalid Page Content" }), { status: 400 });
+		return new Response(JSON.stringify({ message: 'Invalid Page Content' }), { status: 400 });
 	}
 
-	const prompt = `You are an agent that answers a question based on the query. don't say 'based on the context'.\n\n Context:\n${body.query} \nAnswer this question based on the context. Question: ${body.query}\nAnswer:`
+	const prompt = `You are an agent that answers a question based on the query. don't say 'based on the context'.\n\n Context:\n${body.query} \nAnswer this question based on the context. Question: ${body.query}\nAnswer:`;
 	const output = await model.generateContentStream(prompt);
 
 	const response = new Response(
@@ -22,14 +22,14 @@ export async function POST(request: Request, _: CloudflareVectorizeStore, embedd
 				for await (const chunk of output.stream) {
 					const chunkText = await chunk.text();
 					console.log(chunkText);
-					const encodedChunk = converter.encode("data: " + JSON.stringify({ "response": chunkText }) + "\n\n");
+					const encodedChunk = converter.encode('data: ' + JSON.stringify({ response: chunkText }) + '\n\n');
 					controller.enqueue(encodedChunk);
 				}
-				const doneChunk = converter.encode("data: [DONE]");
+				const doneChunk = converter.encode('data: [DONE]');
 				controller.enqueue(doneChunk);
 				controller.close();
-			}
-		})
+			},
+		}),
 	);
 	return response;
 }
diff --git a/apps/cf-ai-backend/src/routes/chat.ts b/apps/cf-ai-backend/src/routes/chat.ts
index 95788d03..75e298b8 100644
--- a/apps/cf-ai-backend/src/routes/chat.ts
+++ b/apps/cf-ai-backend/src/routes/chat.ts
@@ -1,28 +1,28 @@
-import { Content, GenerativeModel } from "@google/generative-ai";
-import { OpenAIEmbeddings } from "../OpenAIEmbedder";
-import { CloudflareVectorizeStore } from "@langchain/cloudflare";
-import { Request } from "@cloudflare/workers-types";
+import { Content, GenerativeModel } from '@google/generative-ai';
+import { OpenAIEmbeddings } from '../OpenAIEmbedder';
+import { CloudflareVectorizeStore } from '@langchain/cloudflare';
+import { Request } from '@cloudflare/workers-types';
 
 export async function POST(request: Request, _: CloudflareVectorizeStore, embeddings: OpenAIEmbeddings, model: GenerativeModel, env?: Env) {
 	const queryparams = new URL(request.url).searchParams;
-	const query = queryparams.get("q");
-	const topK = parseInt(queryparams.get("topK") ?? "5");
-	const user = queryparams.get("user")
-	const spaces = queryparams.get("spaces")
-	const spacesArray = spaces ? spaces.split(",") : undefined
+	const query = queryparams.get('q');
+	const topK = parseInt(queryparams.get('topK') ?? '5');
+	const user = queryparams.get('user');
+	const spaces = queryparams.get('spaces');
+	const spacesArray = spaces ? spaces.split(',') : undefined;
 
-	const sourcesOnly = (queryparams.get("sourcesOnly") ?? "false")
+	const sourcesOnly = queryparams.get('sourcesOnly') ?? 'false';
 
 	if (!user) {
-		return new Response(JSON.stringify({ message: "Invalid User" }), { status: 400 });
+		return new Response(JSON.stringify({ message: 'Invalid User' }), { status: 400 });
 	}
 
 	if (!query) {
-		return new Response(JSON.stringify({ message: "Invalid Query" }), { status: 400 });
+		return new Response(JSON.stringify({ message: 'Invalid Query' }), { status: 400 });
 	}
 	const filter: VectorizeVectorMetadataFilter = {
-		user
-	}
+		user,
+	};
 
 	const responses: VectorizeMatches = { matches: [], count: 0 };
 
@@ -34,12 +34,12 @@ export async function POST(request: Request, _: CloudflareVectorizeStore, embedd
 
 			const resp = await env!.VECTORIZE_INDEX.query(queryAsVector, {
 				topK,
-				filter
+				filter,
 			});
 
 			if (resp.count > 0) {
-				responses.matches.push(...resp.matches)
-				responses.count += resp.count
+				responses.matches.push(...resp.matches);
+				responses.count += resp.count;
 			}
 		}
 	} else {
@@ -47,13 +47,13 @@ export async function POST(request: Request, _: CloudflareVectorizeStore, embedd
 		const resp = await env!.VECTORIZE_INDEX.query(queryAsVector, {
 			topK,
 			filter: {
-				user
-			}
+				user,
+			},
 		});
 
 		if (resp.count > 0) {
-			responses.matches.push(...resp.matches)
-			responses.count += resp.count
+			responses.matches.push(...resp.matches);
+			responses.count += resp.count;
 		}
 	}
 
@@ -61,27 +61,36 @@ export async function POST(request: Request, _: CloudflareVectorizeStore, embedd
 	// 	return new Response(JSON.stringify({ message: "No Results Found" }), { status: 404 });
 	// }
 
-	const highScoreIds = responses.matches.filter(({ score }) => score > 0.35).map(({ id }) => id)
+	const highScoreIds = responses.matches.filter(({ score }) => score > 0.35).map(({ id }) => id);
 
-	if (sourcesOnly === "true") {
+	if (sourcesOnly === 'true') {
 		return new Response(JSON.stringify({ ids: highScoreIds }), { status: 200 });
 	}
 
-	const vec = await env!.VECTORIZE_INDEX.getByIds(highScoreIds)
+	const vec = await env!.VECTORIZE_INDEX.getByIds(highScoreIds);
 
-	const preparedContext = vec.map(({ metadata }) => `Website title: ${metadata!.title}\nDescription: ${metadata!.description}\nURL: ${metadata!.url}\nContent: ${metadata!.text}`).join("\n\n");
+	const preparedContext = vec
+		.map(
+			({ metadata }) =>
+				`Website title: ${metadata!.title}\nDescription: ${metadata!.description}\nURL: ${metadata!.url}\nContent: ${metadata!.text}`,
+		)
+		.join('\n\n');
 
-	const body = await request.json() as {
-		chatHistory?: Content[]
+	const body = (await request.json()) as {
+		chatHistory?: Content[];
 	};
 
 	const defaultHistory = [
 		{
-			role: "user",
-			parts: [{ text: `You are an agent that summarizes a page based on the query. don't say 'based on the context'. I expect you to be like a 'Second Brain'. you will be provided with the context (old saved posts) and questions. Answer accordingly. Answer in markdown format` }],
+			role: 'user',
+			parts: [
+				{
+					text: `You are an agent that summarizes a page based on the query. don't say 'based on the context'. I expect you to be like a 'Second Brain'. you will be provided with the context (old saved posts) and questions. Answer accordingly. Answer in markdown format`,
+				},
+			],
 		},
 		{
-			role: "model",
+			role: 'model',
 			parts: [{ text: "Ok, I am a personal assistant, and will act as a second brain to help with user's queries." }],
 		},
 	] as Content[];
@@ -100,14 +109,14 @@ export async function POST(request: Request, _: CloudflareVectorizeStore, embedd
 				const converter = new TextEncoder();
 				for await (const chunk of output.stream) {
 					const chunkText = await chunk.text();
-					const encodedChunk = converter.encode("data: " + JSON.stringify({ "response": chunkText }) + "\n\n");
+					const encodedChunk = converter.encode('data: ' + JSON.stringify({ response: chunkText }) + '\n\n');
 					controller.enqueue(encodedChunk);
 				}
-				const doneChunk = converter.encode("data: [DONE]");
+				const doneChunk = converter.encode('data: [DONE]');
 				controller.enqueue(doneChunk);
 				controller.close();
-			}
-		})
+			},
+		}),
 	);
 	return response;
 }
diff --git a/apps/cf-ai-backend/src/routes/query.ts b/apps/cf-ai-backend/src/routes/query.ts
index be237d7d..cd5295c5 100644
--- a/apps/cf-ai-backend/src/routes/query.ts
+++ b/apps/cf-ai-backend/src/routes/query.ts
@@ -1,59 +1,65 @@
-import { GenerativeModel } from "@google/generative-ai";
-import { OpenAIEmbeddings } from "../OpenAIEmbedder";
-import { CloudflareVectorizeStore } from "@langchain/cloudflare";
-import { Request } from "@cloudflare/workers-types";
+import { GenerativeModel } from '@google/generative-ai';
+import { OpenAIEmbeddings } from '../OpenAIEmbedder';
+import { CloudflareVectorizeStore } from '@langchain/cloudflare';
+import { Request } from '@cloudflare/workers-types';
 
 export async function GET(request: Request, _: CloudflareVectorizeStore, embeddings: OpenAIEmbeddings, model: GenerativeModel, env?: Env) {
 	const queryparams = new URL(request.url).searchParams;
-	const query = queryparams.get("q");
-	const topK = parseInt(queryparams.get("topK") ?? "5");
-	const user = queryparams.get("user")
-	const space = queryparams.get("space")
+	const query = queryparams.get('q');
+	const topK = parseInt(queryparams.get('topK') ?? '5');
+	const user = queryparams.get('user');
+	const space = queryparams.get('space');
 
-	const sourcesOnly = (queryparams.get("sourcesOnly") ?? "false")
+	const sourcesOnly = queryparams.get('sourcesOnly') ?? 'false';
 
 	if (!user) {
-		return new Response(JSON.stringify({ message: "Invalid User" }), { status: 400 });
+		return new Response(JSON.stringify({ message: 'Invalid User' }), { status: 400 });
 	}
 
 	if (!query) {
-		return new Response(JSON.stringify({ message: "Invalid Query" }), { status: 400 });
+		return new Response(JSON.stringify({ message: 'Invalid Query' }), { status: 400 });
 	}
 
 	const filter: VectorizeVectorMetadataFilter = {
-		user
-	}
+		user,
+	};
 
 	if (space) {
-		filter.space
+		filter.space;
 	}
 
 	const queryAsVector = await embeddings.embedQuery(query);
 
 	const resp = await env!.VECTORIZE_INDEX.query(queryAsVector, {
 		topK,
-		filter
+		filter,
 	});
 
 	if (resp.count === 0) {
-		return new Response(JSON.stringify({ message: "No Results Found" }), { status: 404 });
+		return new Response(JSON.stringify({ message: 'No Results Found' }), { status: 404 });
 	}
 
-	const highScoreIds = resp.matches.filter(({ score }) => score > 0.3).map(({ id }) => id)
+	const highScoreIds = resp.matches.filter(({ score }) => score > 0.3).map(({ id }) => id);
 
-	if (sourcesOnly === "true") {
+	if (sourcesOnly === 'true') {
 		return new Response(JSON.stringify({ ids: highScoreIds }), { status: 200 });
 	}
 
-	const vec = await env!.VECTORIZE_INDEX.getByIds(highScoreIds)
+	const vec = await env!.VECTORIZE_INDEX.getByIds(highScoreIds);
 
 	if (vec.length === 0 || !vec[0].metadata) {
-		return new Response(JSON.stringify({ message: "No Results Found" }), { status: 400 });
+		return new Response(JSON.stringify({ message: 'No Results Found' }), { status: 400 });
 	}
 
-	const preparedContext = vec.slice(0, 3).map(({ metadata }) => `Website title: ${metadata!.title}\nDescription: ${metadata!.description}\nURL: ${metadata!.url}\nContent: ${metadata!.text}`).join("\n\n");
+	const preparedContext = vec
+		.slice(0, 3)
+		.map(
+			({ metadata }) =>
+				`Website title: ${metadata!.title}\nDescription: ${metadata!.description}\nURL: ${metadata!.url}\nContent: ${metadata!.text}`,
+		)
+		.join('\n\n');
 
-	const prompt = `You are an agent that summarizes a page based on the query. Be direct and concise, don't say 'based on the context'.\n\n Context:\n${preparedContext} \nAnswer this question based on the context. Question: ${query}\nAnswer:`
+	const prompt = `You are an agent that summarizes a page based on the query. Be direct and concise, don't say 'based on the context'.\n\n Context:\n${preparedContext} \nAnswer this question based on the context. Question: ${query}\nAnswer:`;
 	const output = await model.generateContentStream(prompt);
 
 	const response = new Response(
@@ -62,14 +68,14 @@ export async function GET(request: Request, _: CloudflareVectorizeStore, embeddi
 				const converter = new TextEncoder();
 				for await (const chunk of output.stream) {
 					const chunkText = await chunk.text();
-					const encodedChunk = converter.encode("data: " + JSON.stringify({ "response": chunkText }) + "\n\n");
+					const encodedChunk = converter.encode('data: ' + JSON.stringify({ response: chunkText }) + '\n\n');
 					controller.enqueue(encodedChunk);
 				}
-				const doneChunk = converter.encode("data: [DONE]");
+				const doneChunk = converter.encode('data: [DONE]');
 				controller.enqueue(doneChunk);
 				controller.close();
-			}
-		})
+			},
+		}),
 	);
 	return response;
 }