Implement MVP Feedbot

app/api/llm-hint/route.ts

@@ -0,0 +1,165 @@
+import { NextRequest, NextResponse } from "next/server";
+import { createClient } from "@/utils/supabase/server";
+import { createClient as createServiceClient } from "@supabase/supabase-js";
+import OpenAI from "openai";
+import * as Sentry from "@sentry/nextjs";
+import { GraderResultTestExtraData } from "@/utils/supabase/DatabaseTypes";
+
+const openai = new OpenAI({
+  apiKey: process.env.OPENAI_API_KEY
+});
+
+export async function POST(request: NextRequest) {
+  try {
+    // Check for required OpenAI configuration
+    if (!process.env.OPENAI_API_KEY) {
+      Sentry.captureMessage("OpenAI API key missing", "error");
+      return NextResponse.json({ error: "OpenAI API key missing" }, { status: 500 });
+    }
+
+    const { testId } = await request.json();
+
+    Sentry.setTag("testId", testId);
+
+    if (!testId || typeof testId !== "number" || testId < 0 || !Number.isInteger(testId)) {
+      return NextResponse.json({ error: "testId must be a non-negative integer" }, { status: 400 });
+    }
+
+    // Verify user has access to this test result
+    const supabase = await createClient();
+    const {
+      data: { user }
+    } = await supabase.auth.getUser();
+
+    if (!user) {
+      return NextResponse.json({ error: "Unauthorized" }, { status: 401 });
+    }
+
+    Sentry.setUser({ id: user.id });
+
+    // Try to get the test result - RLS will handle access control automatically
+    const { data: testResult, error: testError } = await supabase
+      .from("grader_result_tests")
+      .select(
+        `
+        id,
+        extra_data,
+        grader_results!inner (
+          submissions!inner (
+            id
+          )
+        )
+      `
+      )
+      .eq("id", testId)
+      .single();
+
+    if (testError || !testResult) {
+      if (testError) {
+        Sentry.captureException(testError, {
+          tags: { operation: "fetch_test_result", testId: testId.toString() }
+        });
+      }
+      return NextResponse.json({ error: "Test result not found or access denied" }, { status: 404 });
+    }
+
+    // Get the prompt from extra_data and check if hint already exists
+    const extraData = testResult.extra_data as GraderResultTestExtraData | null;
+    if (!extraData?.llm?.prompt) {
+      return NextResponse.json({ error: "No LLM hint prompt found" }, { status: 400 });
+    }
+
+    // Check if hint has already been generated
+    if (extraData.llm.result) {
+      return NextResponse.json({
+        success: true,
+        response: extraData.llm.result,
+        cached: true
+      });
+    }
+
+    // Call OpenAI API with the prompt from the database
+    const apiParams: any = {
+      model: extraData.llm.model || process.env.OPENAI_MODEL || "gpt-4o-mini",
+      messages: [
+        {
+          role: "user",
+          content: extraData.llm.prompt
+        }
+      ],
+      user: `pawtograder:${user.id}`
+    };
+
+    // Only include optional parameters if they're explicitly set in extra_data
+    if (extraData.llm.temperature !== undefined) {
+      const temp = extraData.llm.temperature;
+      if (typeof temp === "number" && temp >= 0 && temp <= 2) {
+        apiParams.temperature = temp;
+      }
+    }
+
+    if (extraData.llm.max_tokens !== undefined) {
+      const maxTokens = extraData.llm.max_tokens;
+      if (typeof maxTokens === "number" && maxTokens > 0) {
+        apiParams.max_completion_tokens = maxTokens;
+      }
+    }
+
+    const completion = await openai.chat.completions.create(apiParams);
+
+    const aiResponse = completion.choices[0]?.message?.content;
+
+    if (!aiResponse) {
+      return NextResponse.json({ error: "No response from AI" }, { status: 500 });
+    }
+
+    // Store the result in the database
+    const updatedExtraData: GraderResultTestExtraData = {
+      ...extraData,
+      llm: {
+        ...extraData.llm,
+        result: aiResponse
+      }
+    };
+
+    // Use service role client for the update since users might not have update permissions
+    const serviceSupabase = createServiceClient(process.env.SUPABASE_URL!, process.env.SUPABASE_SERVICE_ROLE_KEY!);
+
+    const { error: updateError } = await serviceSupabase
+      .from("grader_result_tests")
+      .update({ extra_data: updatedExtraData })
+      .eq("id", testId);
+
+    if (updateError) {
+      Sentry.captureException(updateError, {
+        tags: {
+          operation: "store_llm_hint",
+          testId: testId.toString()
+        },
+        extra: {
+          updateError: updateError.message,
+          testId
+        }
+      });
+      // Still return the result even if storage fails
+    }
+
+    return NextResponse.json({
+      success: true,
+      response: aiResponse,
+      usage: completion.usage,
+      cached: false
+    });
+  } catch (error) {
+    Sentry.captureException(error, {
+      tags: {
+        operation: "llm_hint_api",
+        testId: "unknown"
+      },
+      extra: {
+        error: error instanceof Error ? error.message : String(error)
+      }
+    });
+    return NextResponse.json({ error: "Internal server error" }, { status: 500 });
+  }
+}

app/course/[course_id]/assignments/[assignment_id]/submissions/[submissions_id]/layout.tsx

@@ -40,14 +40,23 @@ import NextLink from "next/link";
 import { useParams, usePathname, useRouter, useSearchParams } from "next/navigation";
 import { ElementType as ReactElementType, useCallback, useEffect, useMemo, useRef, useState } from "react";
 import { BsFileEarmarkCodeFill, BsThreeDots } from "react-icons/bs";
-import { FaBell, FaCheckCircle, FaFile, FaHistory, FaInfo, FaQuestionCircle, FaTimesCircle } from "react-icons/fa";
+import {
+  FaBell,
+  FaCheckCircle,
+  FaFile,
+  FaHistory,
+  FaInfo,
+  FaQuestionCircle,
+  FaRobot,
+  FaTimesCircle
+} from "react-icons/fa";
 import { FiDownloadCloud, FiRepeat, FiSend } from "react-icons/fi";
 import { HiOutlineInformationCircle } from "react-icons/hi";
 import { LuMoon, LuSun } from "react-icons/lu";
 import { PiSignOut } from "react-icons/pi";
 import { RxQuestionMarkCircled } from "react-icons/rx";
 import { TbMathFunction } from "react-icons/tb";
-import { GraderResultTestData } from "./results/page";
+import { GraderResultTestExtraData } from "@/utils/supabase/DatabaseTypes";
 import { linkToSubPage } from "./utils";
 import { useAssignmentController } from "@/hooks/useAssignment";
 
@@ -429,8 +438,10 @@ function TestResults() {
       </Heading>
       {testResults?.map((test) => {
         let icon;
-        const extraData = test.extra_data as GraderResultTestData;
-        if (extraData?.icon && iconMap[extraData.icon]) {
+        const extraData = test.extra_data as GraderResultTestExtraData;
+        if (extraData?.llm?.prompt || extraData?.llm?.result) {
+          icon = <Icon as={FaRobot} />;
+        } else if (extraData?.icon && iconMap[extraData.icon]) {
           icon = <Icon as={iconMap[extraData.icon]} />;
         } else if (test.score === 0 && test.max_score === 0) {
           icon = <Icon as={FaInfo} color="fg.info" />;