Minor fixes to AI docs (#689)

* Minor fixes to AI docs

* Restore ai folders introductions

* Implement requested changes

* Add agent setup tutorial & workshop as company resources

* Fix workshop youtube URL

* Re-generate code examples

components/ResourceHub/company-resources.json

@@ -182,5 +182,33 @@
     ],
     "image": "https://docs.safe.global/og_image.png",
     "lastChecked": "2025-01-21"
+  },
+  {
+    "name": "Create an AI agent that can interact with your Safe",
+    "url": "https://docs.safe.global/home/ai-agent-setup",
+    "type": "Blog Post",
+    "date": "2025-02-11",
+    "description": "In this tutorial, we will learn how to set up and deploy an AI agent that has capabilities to access a Safe and prepare transactions for it 🤖.",
+    "tags": [
+      "Tutorial",
+      "Safe Core SDK",
+      "AI"
+    ],
+    "lastChecked": "2025-02-13"
+  },
+  {
+    "name": "Workshop: Build reliable AI agents with Safe capabilities",
+    "url": "https://www.youtube.com/watch?v=y90oH7w8h8U",
+    "type": "Video",
+    "date": "2025-02-11",
+    "description": "In this Webinar, Safe developer Louis will provide an introduction into how to setup your AI agent with a Safe Smart Account.",
+    "tags": [
+      "Tutorial",
+      "Safe Core SDK",
+      "AI"
+    ],
+    "lastChecked": "2025-02-13"
   }
 ]
+
+

examples/ai-agent/tools/safe.ts

@@ -1,6 +1,6 @@
 import { z } from "zod";
 import Safe from "@safe-global/protocol-kit";
-import { createPublicClient, http } from "viem";
+import { createPublicClient, formatEther, http } from "viem";
 import { sepolia } from "viem/chains";
 
 export const getEthBalance = async ({ address, chainId }) => {
@@ -25,11 +25,9 @@ export const getEthBalance = async ({ address, chainId }) => {
   const weiBalance = ethBalanceData.find(
     (element) => element?.tokenAddress === null && element?.token === null
   )?.balance;
-  const ethBalance = BigInt(weiBalance) / BigInt(10 ** 18); // Convert from wei to eth
+  const ethBalance = formatEther(weiBalance); // Convert from wei to eth
 
-  return `The current balance of the Safe Multisig at address ${address} is ${ethBalance.toLocaleString(
-    "en-US"
-  )} ETH.`;
+  return `The current balance of the Safe Multisig at address ${address} is ${ethBalance} ETH.`;
 };
 
 export const deployNewSafe = async () => {

pages/home/ai-agent-quickstarts/basic-agent-setup.mdx

@@ -28,9 +28,11 @@ const RPC_URL = 'https://rpc.ankr.com/eth_sepolia'
 const safeClient = await Safe.init({
   provider: RPC_URL,
   signer: SIGNER_PRIVATE_KEY,
-  safeOptions: {
-    owners: [SIGNER_ADDRESS],
-    threshold: 1
+  predictedSafe: {
+    safeAccountConfig: {
+      owners: [SIGNER_ADDRESS],
+      threshold: 1
+    }
   }
 })
 ```

pages/home/ai-agent-quickstarts/introduction.mdx

@@ -7,4 +7,4 @@ We describe these different setups:
 1. [Basic Agent Setup](./basic-agent-setup.mdx): The simplest and fastest setup for your agent.
 2. [Human Approval for agent actions](./human-approval.mdx): The AI agent proposes transactions, human signers approve and execute the transaction. Highest security, but not the fastest.
 3. [Multi Agent Setup](./multi-agent-setup.mdx): Multiple Agents are signers of a Smart Account. They have to approve and execute each others transactions.
-4. [Agent with a spending limit](./agent-with-spending-limit.mdx): An AI agent gets a spending limit on your treasury or DAO funds. This advanced setup allows fast transactions with a good amount of security.
+4. [Agent with a spending limit](./agent-with-spending-limit.mdx): An AI agent gets a spending limit on your treasury or DAO funds. This advanced setup allows fast transactions with a good amount of security.

pages/home/ai-agent-setup.mdx

@@ -2,8 +2,6 @@ import { Callout } from 'nextra/components'
 
 # Create an AI agent that can interact with your Safe
 
-At Safe, we envision a future where an important part of the onchain economy is driven by AI agents. We support this vision by organizing one of the [biggest AI x Crypto hackathons to date](https://agentathon.ai), as well as providing concrete [guides](./ai-agent-quickstarts/introduction) to help you get started.
-
 In this tutorial, we will learn how to set up and deploy an AI agent that has capabilities to access a Safe and prepare transactions for it 🤖. We will use [LangChain](https://www.langchain.com/agents) to set up the agent, the [Protocol Kit](../sdk/protocol-kit.mdx) to interact with Safes, and [`ollama`](https://ollama.com/) to load and run the agents locally.
 
 **Note:** If you wish to follow along using the completed project, you can [check out the GitHub repository](https://github.com/5afe/safe-ai-agent-tutorial) for this tutorial.