[{"_1":2,"_170":-5,"_171":-5},"loaderData",{"_3":4,"_91":92},"root",{"_5":6,"_34":35,"_64":65},"ctx",{"_7":8,"_23":-5,"_24":25,"_26":27,"_28":20,"_29":20,"_30":20,"_31":32,"_33":20},"auth",{"_9":-5,"_10":-5,"_11":-5,"_12":-5,"_13":-5,"_14":-5,"_15":-5,"_16":-5,"_17":-5,"_18":-5,"_19":20,"_21":22},"sessionClaims","sessionId","sessionStatus","userId","actor","orgId","orgRole","orgSlug","orgPermissions","factorVerificationAge","isAuthenticated",false,"tokenType","session_token","user","preferencesKeybindings","default","preferencesTheme","light","preferencesFixers","preferencesCopilot","preferencesSelectionAi","notificationCount",0,"flagDenols","ENV",{"_36":37,"_38":39,"_40":41,"_42":43,"_44":45,"_46":47,"_48":49,"_50":51,"_52":53,"_54":55,"_56":57,"_58":59,"_60":61,"_62":63},"REACT_APP_SENTRY_DSN","https://586020d86f05489db63a554690acf1e9@o1398359.ingest.sentry.io/6724499","VAL_RUNTIME_NAME","prod","RENDER_GIT_COMMIT","a31a1550ad486cf340b818a53bc6ddc8c7db0295","RENDER_GIT_BRANCH","main","CLERK_PUBLISHABLE_KEY","pk_live_Y2xlcmsudmFsLnRvd24k","POSTHOG_API_TOKEN","phc_nmoyuAkStEZuUp69rYlg1DZgrIt19Gx0Y4Hwf8tGgJP","POSTHOG_API_HOST","https://wart.val.town","WEB_API_TEMPLATE","https://{val}.web.val.run","APP_URL","https://www.val.town","MODULE_URL","https://esm.town","VANITY_WEB_API_TEMPLATE","https://{label}.val.run/","DENOLS_URL","https://lsp.val.town","NODE_ENV","production","API_URL","https://api.val.town","clerkState",{"_66":67},"__internal_clerk_state",{"_68":69,"_70":45,"_71":72,"_73":72,"_74":20,"_75":72,"_76":72,"_77":72,"_78":72,"_79":80,"_89":20,"_90":20},"__clerk_ssr_state",{"_21":22,"_9":-5,"_10":-5,"_11":-5,"_12":-5,"_13":-5,"_14":-5,"_15":-5,"_16":-5,"_17":-5,"_18":-5,"_19":20},"__publishableKey","__proxyUrl","","__domain","__isSatellite","__signInUrl","__signUpUrl","__afterSignInUrl","__afterSignUpUrl","__clerk_debug",{"_81":20,"_19":20,"_82":72,"_83":84,"_85":72,"_86":45,"_87":20,"_88":72},"isSignedIn","proxyUrl","reason","session-token-and-uat-missing","message","publishableKey","isSatellite","domain","__telemetryDisabled","__telemetryDebug","routes/embed.x.$handle.$name.$",{"_93":94,"_95":96,"_97":98,"_99":100,"_149":150,"_167":168,"_169":-5},"handle","veltsonbastien","name","dropbox-demo-endpoint","path","main.ts","project",{"_101":102,"_95":96,"_103":104,"_105":-5,"_106":20,"_107":108,"_109":-5,"_110":-5,"_12":111,"_112":113,"_114":115,"_116":115,"_117":-5,"_118":32,"_14":119,"_120":121,"_122":121,"_123":124,"_125":124,"_126":124,"_127":121,"_128":121,"_23":129,"_132":133,"_135":136,"_148":20},"id","6830e180-d097-11f0-a68c-42dde27851f2","createdAt",["D",1764800915512],"deletedAt","legacyEnvVars","privacy","public","description","imageUrl","019a5bff-0bfc-72ad-9aef-d0d797d23edb","pinnedFileId","68332f76-d097-11f0-a68c-42dde27851f2","pinPreview",true,"pinMigrated","screenshotVersion","likeCount","019a5bff-0bfc-72ad-9aef-d7932d5b41c9","userPerm","r","valPerm","blobPerm","rw","emailPerm","sqlitePerm","projectPerm","telemetryPerm",{"_93":94,"_130":131},"profileImageUrl","https://img.clerk.com/eyJ0eXBlIjoicHJveHkiLCJzcmMiOiJodHRwczovL2ltYWdlcy5jbGVyay5kZXYvb2F1dGhfZ2l0aHViL2ltZ18zNTg3cEJMaHZOMFlRVVE1SmczVWx1a3lLRFYifQ","org",{"_93":94,"_134":131,"_12":111},"avatar","branch",{"_137":138,"_95":43,"_101":139,"_140":102,"_141":-5,"_142":-5,"_143":-5,"_144":-5,"_145":-5,"_146":96,"_147":94},"version",5,"68312c44-d097-11f0-a68c-42dde27851f2","projectId","forkedBranchId","forkedBranchVersion","forkedProjectPrivacy","forkedProjectId","forkedProjectOrgId","projectName","userHandle","isOwned","file",{"_101":151,"_95":98,"_152":113,"_153":-5,"_154":139,"_155":138,"_156":115,"_103":157,"_137":158,"_159":160,"_161":162,"_163":20,"_164":165,"_97":166},"068da10e-d176-11f0-a720-42dde27851f2","fileId","parentId","branchId","branchVersion","latest",["D",1764896529468],4,"fileType","http","content","// Val.town RAG Endpoint\n// POST /rag with JSON body: { \"question\": \"your question here\" }\n// Required header: x-api-key: \n\nimport { OpenAI } from \"npm:openai@4\";\nimport { Pinecone } from \"npm:@pinecone-database/pinecone@3\";\nimport Anthropic from \"npm:@anthropic-ai/sdk@0.27\";\n\n// Env vars\nconst AUTH_TOKEN = Deno.env.get(\"AUTH_TOKEN\");\nconst OPENAI_API_KEY = Deno.env.get(\"OPENAI_API_KEY\");\nconst PINECONE_API_KEY = Deno.env.get(\"PINECONE_API_KEY\");\nconst PINECONE_INDEX_NAME = Deno.env.get(\"PINECONE_INDEX_NAME\") ||\n \"aws-ring-embeddings-small\";\nconst PINECONE_DIMENSIONS = parseInt(\n Deno.env.get(\"PINECONE_DIMENSIONS\") || \"1536\",\n);\nconst OPENAI_EMBEDDINGS_MODEL = Deno.env.get(\"OPENAI_EMBEDDINGS_MODEL\") ||\n \"text-embedding-3-small\";\nconst ANTHROPIC_API_KEY = Deno.env.get(\"ANTHROPIC_API_KEY\");\n\n// Validate required environment variables\nif (!OPENAI_API_KEY) {\n throw new Error(\"OPENAI_API_KEY environment variable is required\");\n}\nif (!PINECONE_API_KEY) {\n throw new Error(\"PINECONE_API_KEY environment variable is required\");\n}\nif (!ANTHROPIC_API_KEY) {\n throw new Error(\"ANTHROPIC_API_KEY environment variable is required\");\n}\nif (!AUTH_TOKEN) {\n throw new Error(\"AUTH_TOKEN environment variable is required\");\n}\n\n// Clients (reused across invocations)\nconst openai = new OpenAI({ apiKey: OPENAI_API_KEY });\nconst pinecone = new Pinecone({ apiKey: PINECONE_API_KEY });\nconst anthropic = new Anthropic({ apiKey: ANTHROPIC_API_KEY });\n\nconst index = pinecone.index(PINECONE_INDEX_NAME);\n\n// --- helpers -------------------------------------------------------\n\nasync function embedQuestion(question: string): Promise {\n const response = await openai.embeddings.create({\n model: OPENAI_EMBEDDINGS_MODEL,\n input: question,\n dimensions: PINECONE_DIMENSIONS,\n });\n return response.data[0].embedding;\n}\n\nasync function queryDatabase(embedding: number[]): Promise {\n const maxAttempts = 3;\n\n for (let attempt = 1; attempt <= maxAttempts; attempt++) {\n try {\n const results = await index.query({\n vector: embedding,\n topK: 2,\n includeMetadata: true,\n });\n\n const contexts = results.matches\n .map((match: any) => {\n const question = match.metadata?.question || \"\";\n const answer = match.metadata?.answer || \"\";\n if (question && answer) {\n return `Question: ${question}\\nAnswer: ${answer}`;\n }\n return \"\";\n })\n .filter((content: string) => content.length > 0);\n\n return contexts.join(\"\\n\\n\");\n } catch (err: any) {\n console.error(\n `[RAG] Pinecone query error on attempt ${attempt}:`,\n err?.message,\n );\n if (attempt === maxAttempts) throw err;\n await new Promise((r) => setTimeout(r, 200 * attempt)); // simple backoff\n }\n }\n\n return \"\";\n}\n\nasync function generateResponse(\n question: string,\n context: string,\n): Promise {\n const systemPrompt = `**Generate Response to User Query**\n**Step 1: Parse Context Information**\nExtract and utilize relevant knowledge from the provided context within \\`\\` XML tags.\nThe context contains context that is possibly relevant to the user's query.\n\n**Step 2: Analyze User Query**\nCarefully read and comprehend the user's query and intent behind the question.\n\n**Step 3: Determine Response**\nIf the answer to the user's query can be directly inferred from the context information, provide a concise and accurate response matching as closely to the context as possible.\nIf the answer is not found in the context, return \"I'm sorry, but unfortunately I don't have the answer to that question. I can transfer to someone who can better help you with that\".\n\n**Step 4: Generate Response**\nDo not repeat the question in the response. The response should match the format of the context as closely as possible.`;\n\n const userPrompt = `**Generate Response to User Query**\nUser Query: ${question}\n\n\n${context}\n`;\n\n const response = await anthropic.messages.create({\n model: \"claude-haiku-4-5\",\n max_tokens: 300,\n system: systemPrompt,\n messages: [\n {\n role: \"user\",\n content: userPrompt,\n },\n ],\n });\n\n return response.content[0].type === \"text\" ? response.content[0].text : \"\";\n}\n\n// generic timer wrapper\nasync function timeStep(\n label: string,\n fn: () => Promise,\n): Promise<{ result: T; ms: number }> {\n const start = performance.now();\n const result = await fn();\n const ms = performance.now() - start;\n console.log(`[RAG] ${label} ms:`, ms.toFixed(2));\n return { result, ms };\n}\n\n// --- main handler --------------------------------------------------\n\nexport default async function (req: Request): Promise {\n const startTime = performance.now();\n\n // timings we’ll fill as we go\n const timings: {\n total_seconds?: number;\n embed_ms?: number;\n pinecone_query_ms?: number;\n anthropic_ms?: number;\n } = {};\n\n try {\n // Auth\n const apiKey = req.headers.get(\"x-api-key\");\n if (!apiKey) {\n return Response.json(\n { error: \"Missing API key. Please provide x-api-key header\" },\n { status: 401 },\n );\n }\n if (apiKey !== AUTH_TOKEN) {\n return Response.json(\n { error: \"Invalid API key\" },\n { status: 401 },\n );\n }\n\n // Method check\n if (req.method !== \"POST\") {\n return Response.json(\n { error: \"Only POST requests are accepted\" },\n { status: 405 },\n );\n }\n\n // Parse body\n const body = await req.json();\n const question = body.question;\n\n if (!question || typeof question !== \"string\") {\n return Response.json(\n { error: \"Question is required and must be a string\" },\n { status: 400 },\n );\n }\n\n // Step 1: embeddings\n const { result: embedding, ms: embedMs } = await timeStep(\n \"embedQuestion\",\n () => embedQuestion(question),\n );\n timings.embed_ms = embedMs;\n\n // Step 2: Pinecone\n const { result: context, ms: pineconeMs } = await timeStep(\n \"queryDatabase\",\n () => queryDatabase(embedding),\n );\n timings.pinecone_query_ms = pineconeMs;\n\n // Step 3: Anthropic\n const { result: answer, ms: anthropicMs } = await timeStep(\n \"generateResponse\",\n () => generateResponse(question, context),\n );\n timings.anthropic_ms = anthropicMs;\n\n // Total\n const totalSeconds = (performance.now() - startTime) / 1000;\n timings.total_seconds = parseFloat(totalSeconds.toFixed(3));\n\n return Response.json({\n question,\n answer,\n context_length: context.length,\n timings,\n });\n } catch (error: any) {\n const totalSeconds = (performance.now() - startTime) / 1000;\n timings.total_seconds = parseFloat(totalSeconds.toFixed(3));\n\n console.error(\n `[RAG] Request failed after ${timings.total_seconds}s:`,\n error,\n );\n\n return Response.json(\n {\n error: error?.message || \"Internal server error\",\n timings,\n },\n { status: 500 },\n );\n }\n}","deleted","label","vapi-dropbox-demo",[98],"permalink","https://www.val.town/x/veltsonbastien/dropbox-demo-endpoint/code/main.ts","extension","actionData","errors"]