ComposioHQ · sohamganatra · Oct 9, 2024 · Oct 8, 2024 · Oct 8, 2024 · Oct 8, 2024
@@ -0,0 +1,90 @@
+import dotenv from 'dotenv';
+import { ChatOpenAI } from "@langchain/openai";
+import { AgentExecutor, createOpenAIFunctionsAgent } from "langchain/agents";
+import { pull } from "langchain/hub";
+import { LangchainToolSet } from "composio-core";
+import readline from 'readline';
+
+dotenv.config();
+// Access environment variables
+const openaiapiKey = process.env.OPENAI_API_KEY;
+const composioapiKey = process.env.COMPOSIO_API_KEY;
+
+// Use environment variables in your code
+// ...
+
+// Initialize the language model
+const llm = new ChatOpenAI({ model: "gpt-4-turbo", apiKey: openaiapiKey});
+
+// Define tools for the agents
+const composioToolset = new LangchainToolSet({
+    apiKey: composioapiKey
+});
+
+//Sample input code, change according to code you want to debug
+const input = `
+<!doctype html>
+<html>
+<head>
+  <meta charset="utf-8">
+  <title>Untitled Document</title>
+</head>
+<body>
+  <button type="button" onClick="myfunc()"></button>
+  <script>
+    function myfunc() {
+      var a = prompt("Enter the year?");
+      if (a % 4 == 0) {
+        if (a % 100 == 0) {
+          if (a % 400 == 0) {
+            window.alert("it is a leap year");
+          }
+          else:
+            window.alert("it is not a leap year");
+        }
+        else:
+          window.alert("it is a leap year");
+      }
+    }
+  </script>
+</body>
+</html>
+
+`;
+
+
+// Setup Todo
+const todo = `
+  This is the code given to you ${input}.
+  Execute it and find all the errors and bugs within it, then find solutions for these
+  either by trying various modifications or going on internet. Suggest links that can be 
+  read up on if you cant find a concrete solution.  
+`;
+
+async function runAgent() {
+    const tools = await composioToolset.getTools({
+        apps:['codeinterpreter','exa','tavily']
+    });
+
+    const prompt = await pull("hwchase17/openai-functions-agent");
+    const agent = await createOpenAIFunctionsAgent({
+        llm,
+        tools,
+        prompt
+    });
+
+    const agentExecutor = new AgentExecutor({
+        agent,
+        tools,
+        verbose: true,
+    });
+
+    const result = await agentExecutor.invoke({
+        input: todo
+    });
+
+    console.log(result.output);
+    return "Agent execution completed";
+}
+
+runAgent().then(console.log).catch(console.error);