### Start Stdio Transport with Inspector

Source: https://github.com/pipedreamhq/pipedream/blob/master/modelcontextprotocol/README.md

Use this command to run the Stdio transport example with the MCP Inspector. This setup has only been tested with the MCP Inspector.

```bash
npx @modelcontextprotocol/inspector bun src/stdio.ts
```

--------------------------------

### Install Dependencies with npm

Source: https://github.com/pipedreamhq/pipedream/blob/master/types/README.md

Use this command to install project dependencies. It is required before running tests or building the project.

```bash
npm install
```

--------------------------------

### Create a Demo Action

Source: https://github.com/pipedreamhq/pipedream/blob/master/README.md

This action accepts a 'name' as input and returns a greeting. It's useful for demonstrating custom action creation.

```javascript
export default {
  name: "Action Demo",
  description: "This is a demo action",
  key: "action_demo",
  version: "0.0.1",
  type: "action",
  props: {
    name: {
      type: "string",
      label: "Name",
    },
  },
  async run() {
    return `hello ${this.name}!`;
  },
};

```

--------------------------------

### AWS IAM Policy for CloudWatch Logs Insights

Source: https://github.com/pipedreamhq/pipedream/blob/master/components/aws/sources/new-records-returned-by-cloudwatch-logs-insights-query/README.md

This IAM policy grants the necessary permissions to start, describe, get, and stop CloudWatch Logs Insights queries. Ensure your IAM user has at least these permissions.

```json
{
  "Version": "2012-10-17",
  "Statement": [
    {
      "Sid": "VisualEditor0",
      "Effect": "Allow",
      "Action": "logs:StartQuery",
      "Resource": "arn:aws:logs:*:*:log-group:*"
    },
    {
      "Sid": "VisualEditor1",
      "Effect": "Allow",
      "Action": "logs:DescribeLogGroups",
      "Resource": "arn:aws:logs:*:*:log-group:*"
    },
    {
      "Sid": "VisualEditor2",
      "Effect": "Allow",
      "Action": ["logs:GetQueryResults", "logs:StopQuery"],
      "Resource": "*"
    }
  ]
}
```

--------------------------------

### Basic OpenAI Tool Integration with @pipedream/ai

Source: https://github.com/pipedreamhq/pipedream/blob/master/packages/ai/README.md

This example demonstrates how to initialize and use the OpenAiTools class to fetch tools for a specific application (e.g., Slack) and then handle an OpenAI completion that utilizes these tools. Ensure you replace '<add user id here>' with a unique identifier for your user.

```typescript
import { OpenAiTools } from "@pipedream/ai"
import { OpenAI } from "openai"

const openai = new OpenAI()

// Replace with a unique identifier for your user
const userId = <add user id here>

const openAiTools = new OpenAiTools(userId)
const tools = await openAiTools.getTools({
  app: "slack",
})

const completion = await openai.chat.completions.create({
  messages: [
    {
      role: "user",
      content: "Send a joke to #random channel",
    },
  ],
  model: "gpt-4o",
  tools,
})

const results = await openAiTools.handleCompletion(completion)

console.log(JSON.stringify(results, null, 2))

```

--------------------------------

### Build Project with npm

Source: https://github.com/pipedreamhq/pipedream/blob/master/types/README.md

Run this command to build the project. This is typically done after making code changes or before publishing.

```bash
npm run build
```