### Simple General.AI Quick Start

Source: https://github.com/nixaut-codelabs/general.ai/blob/main/README.md

This snippet demonstrates how to initialize GeneralAI with an OpenAI client and generate a simple response using the agent. It shows basic setup and a common use case for generating text.

```typescript
import OpenAI from "openai";
import { GeneralAI } from "@lightining/general.ai";

const openai = new OpenAI({
  apiKey: process.env.OPENAI_API_KEY,
});

const generalAI = new GeneralAI({ openai });

const result = await generalAI.agent.generate({
  endpoint: "chat_completions",
  model: "gpt-5.4-mini",
  messages: [
    { role: "user", content: "Say hello briefly in Turkish." },
  ],
});

console.log(result.cleaned);
```

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

### Install General.AI with Bun

Source: https://github.com/nixaut-codelabs/general.ai/blob/main/README.md

Install the General.AI package and the 'openai' package using Bun. This is the recommended installation method for Bun environments.

```bash
bun add @lightining/general.ai openai
```

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

### Install General.AI with npm

Source: https://github.com/nixaut-codelabs/general.ai/blob/main/README.md

Install the General.AI package and the 'openai' package using npm. This is the standard installation method for Node.js environments.

```bash
npm install @lightining/general.ai openai
```

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

### Install Beta Version of General.AI with Bun

Source: https://github.com/nixaut-codelabs/general.ai/blob/main/README.md

Install the beta version of the General.AI package using Bun to access the latest runtime features and capabilities. This is for users who want to test the newest developments in a Bun environment.

```bash
bun add @lightining/general.ai@beta openai
```

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

### Install Beta Version of General.AI with npm

Source: https://github.com/nixaut-codelabs/general.ai/blob/main/README.md

Install the beta version of the General.AI package using npm to access the latest runtime features and capabilities. This is for users who want to test the newest developments.

```bash
npm install @lightining/general.ai@beta openai
```

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

### Render Prompts with Task Context Override

Source: https://github.com/nixaut-codelabs/general.ai/blob/main/README.md

Example of rendering prompts using the General.AI SDK, specifically overriding the task section with block content.

```typescript
const prompt = await generalAI.agent.renderPrompts({
  endpoint: "responses",
  model: "gpt-5.4-mini",
  messages: [{ role: "user", content: "Hello" }],
  prompts: {
    sections: {
      task: "Task override.\n{block:task_context}",
    },
  },
});
```

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

### Advanced Agent Generation with Tools and Subagents

Source: https://github.com/nixaut-codelabs/general.ai/blob/main/README.md

Demonstrates an advanced agent generation call that utilizes tools, subagents, and context management. This example showcases General.AI's orchestration capabilities.

```typescript
const result = await generalAI.agent.generate({
  endpoint: "chat_completions",
  model: "gpt-5.4-mini",
  messages: [
    {
      role: "user",
      content: "Use tools if needed, delegate arithmetic to a subagent if useful, and give me a short final answer.",
    },
  ],
  compatibility: {
    profile: "classic_v2",
  },
  tools: {
    registry: [weatherTool, calculatorTool],
  },
  subagents: {
    registry: [mathHelper],
  },
  context: {
    mode: "auto",
    strategy: "hybrid",
  },
});

console.log(result.cleaned);
console.log(result.meta.contextOperations);
console.log(result.meta.warnings);
```

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

### Initialize OpenAI Client with General.AI

Source: https://github.com/nixaut-codelabs/general.ai/blob/main/README.md

Demonstrates how to initialize the GeneralAI client using an existing OpenAI client. Ensure your OPENAI_API_KEY is set in the environment.

```typescript
import OpenAI from "openai";
import { GeneralAI } from "@lightining/general.ai";

const openai = new OpenAI({
  apiKey: process.env.OPENAI_API_KEY,
});

const generalAI = new GeneralAI({ openai });

const result = await generalAI.agent.generate({
  endpoint: "chat_completions",
  model: "gpt-5.4-mini",
  preset: "classic_safe",
  intelligence: "high",
  messages: [
    { role: "user", content: "Say hello briefly in Turkish." },
  ],
});

console.log(result.cleaned);
console.log(result.meta.warnings);
console.log(result.meta.performance.speed);
console.log(result.meta.configuration);
```

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

### Initialize General.AI with Provider Configuration

Source: https://github.com/nixaut-codelabs/general.ai/blob/main/README.md

Shows how to initialize GeneralAI by directly configuring the provider, including API keys, rotation strategies, and queue settings. Ensure NVIDIA_KEY_A and NVIDIA_KEY_B are set in the environment.

```typescript
import { GeneralAI } from "@lightining/general.ai";

const generalAI = new GeneralAI({
  provider: {
    baseURL: "https://integrate.api.nvidia.com/v1",
    apiKeys: [
      process.env.NVIDIA_KEY_A!,
      process.env.NVIDIA_KEY_B!,
    ],
    rotation: {
      strategy: "round_robin",
      onRateLimit: "next_key",
      maxRateLimitHandoffs: 2,
    },
    queue: {
      maxConcurrentRequests: 3,
    },
  },
});
```

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

### Run Manual Test Script

Source: https://github.com/nixaut-codelabs/general.ai/blob/main/README.md

Execute the manual test script using bun. This script can also exercise optional live provider checks.

```bash
bun run test.js
```

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

### Configure Provider Pool and Queue

Source: https://github.com/nixaut-codelabs/general.ai/blob/main/README.md

Initialize GeneralAI with a provider configuration, including API keys, rotation strategy, and queue management. Ensure environment variables for API keys are set.

```typescript
const generalAI = new GeneralAI({
  provider: {
    name: "nvidia",
    baseURL: "https://integrate.api.nvidia.com/v1",
    apiKeys: [
      { key: process.env.NVIDIA_KEY_A!, label: "nvidia-a" },
      { key: process.env.NVIDIA_KEY_B!, label: "nvidia-b" },
      { key: process.env.NVIDIA_KEY_C!, label: "nvidia-c" },
    ],
    rotation: {
      strategy: "round_robin",
      onRateLimit: "next_key",
      maxRateLimitHandoffs: 3,
      revisitKeysInSameRequest: false,
    },
    queue: {
      enabled: true,
      maxConcurrentRequests: 3,
      maxQueuedRequests: 100,
      strategy: "fifo",
    },
  },
});
```

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

### Apply Raw Prompt Overrides

Source: https://github.com/nixaut-codelabs/general.ai/blob/main/README.md

Demonstrates how to apply raw prompt overrides by prepending, appending, or replacing the entire rendered prompt content.

```typescript
prompts: {
  raw: {
    prepend: "Extra preamble",
    append: "Extra appendix",
    replace: "Replace the entire rendered prompt",
  },
}
```

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

### Run Default Tests

Source: https://github.com/nixaut-codelabs/general.ai/blob/main/README.md

Execute deterministic tests using npm. This command runs the standard test suite for the project.

```bash
npm test
```

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

### Native SDK: Create Response

Source: https://github.com/nixaut-codelabs/general.ai/blob/main/README.md

Uses the native General.AI interface to create a response, mimicking OpenAI SDK behavior. This is suitable when exact OpenAI compatibility is required.

```typescript
const response = await generalAI.native.responses.create({
  model: "gpt-5.4-mini",
  input: "Give a one-sentence explanation of prompt caching.",
});
```

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

### Set Environment Variables for Live Provider Checks

Source: https://github.com/nixaut-codelabs/general.ai/blob/main/README.md

Configure environment variables to enable live provider checks in manual test scripts. Set API key, base URL, and model, or skip live checks.

```bash
GENERAL_AI_API_KEY=...
GENERAL_AI_BASE_URL=...
GENERAL_AI_MODEL=...
GENERAL_AI_SKIP_LIVE=1
```

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

### Set Compatibility Profile

Source: https://github.com/nixaut-codelabs/general.ai/blob/main/README.md

Configure the compatibility profile for OpenAI-compatible providers. Use 'classic_v2' for stricter gateways.

```typescript
compatibility: {
  profile: "classic_v2",
}
```

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

### Define a Simple Echo Tool

Source: https://github.com/nixaut-codelabs/general.ai/blob/main/README.md

Define a runtime tool using 'defineTool' for testing purposes. The input schema specifies the expected arguments for the tool's execution.

```typescript
import { defineTool } from "@lightining/general.ai";

const echoTool = defineTool({
  name: "echo",
  description: "Echo text back for runtime testing.",
  inputSchema: {
    type: "object",
    additionalProperties: false,
    properties: {
      text: { type: "string" },
    },
    required: ["text"],
  },
  async execute(args) {
    return { echoed: args.text };
  },
});
```

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

### Native SDK: Create Chat Completion

Source: https://github.com/nixaut-codelabs/general.ai/blob/main/README.md

Uses the native General.AI interface to create a chat completion, mirroring OpenAI's chat completions API. This is useful for direct OpenAI SDK compatibility.

```typescript
const completion = await generalAI.native.chat.completions.create({
  model: "gpt-5.4-mini",
  messages: [{"role": "user", "content": "Say hello in one sentence." }],
});
```

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

### Run Cross-Runtime Smoke Tests

Source: https://github.com/nixaut-codelabs/general.ai/blob/main/README.md

Execute smoke tests across different runtimes using npm. This verifies basic functionality in various environments.

```bash
npm run smoke
```

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

### Configure Context Management

Source: https://github.com/nixaut-codelabs/general.ai/blob/main/README.md

Enable context management and specify the mode and strategy for handling conversational context. Supports various strategies like summarization and dropping old messages.

```typescript
context: {
  enabled: true,
  mode: "auto",
  strategy: "hybrid",
  trigger: {
    contextRatio: 0.9,
  },
}
```

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

### Configure Thinking Parameters

Source: https://github.com/nixaut-codelabs/general.ai/blob/main/README.md

Configure the thinking parameters for agent execution, including mode, strategy, and effort. Ensure 'enabled' is set to true to activate thinking features.

```typescript
thinking: {
  enabled: true,
  mode: "hybrid",
  strategy: "checkpointed",
  checkpointFormat: "structured",
  effort: "high",
}
```

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

### Generate with Preset and Intelligence

Source: https://github.com/nixaut-codelabs/general.ai/blob/main/README.md

Generate content using specific presets and intelligence levels for runtime posture and prompt guidance. Ensure the 'messages' array is correctly formatted.

```typescript
const result = await generalAI.agent.generate({
  endpoint: "chat_completions",
  model: "gpt-5.4-mini",
  preset: "agentic",
  intelligence: "high",
  messages: [{ role: "user", content: "Solve this carefully." }],
});
```

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

### Define a Root-Only Tool

Source: https://github.com/nixaut-codelabs/general.ai/blob/main/README.md

Define a tool that is only accessible from the root agent by setting 'subagents: false' in the access configuration.

```typescript
const rootOnlyTool = defineTool({
  name: "root_only",
  description: "Only callable from the root agent.",
  access: {
    subagents: false,
  },
  async execute() {
    return { ok: true };
  },
});
```

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

### Configure Safety Settings

Source: https://github.com/nixaut-codelabs/general.ai/blob/main/README.md

Enable or disable safety features and set modes for input and output moderation within the agent protocol.

```typescript
safety: {
  enabled: true,
  mode: "balanced",
  input: {
    enabled: true,
  },
  output: {
    enabled: true,
  },
}
```

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

### Stream Agent Events

Source: https://github.com/nixaut-codelabs/general.ai/blob/main/README.md

Initiate an agent stream to receive parsed runtime events and cleaned writing deltas. Process events like 'writing_delta' for real-time output.

```typescript
const stream = generalAI.agent.stream({
  endpoint: "responses",
  model: "gpt-5.4-mini",
  messages: [{ role: "user", content: "Say hello." }],
});

for await (const event of stream) {
  if (event.type === "writing_delta") {
    process.stdout.write(event.text);
  }
}
```

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

### Define a Subagent

Source: https://github.com/nixaut-codelabs/general.ai/blob/main/README.md

Define a subagent with its own configuration, including model, instructions, and request parameters. Subagents can override many root-level configurations.

```typescript
import { defineSubagent } from "@lightining/general.ai";

const mathHelper = defineSubagent({
  name: "math_helper",
  description: "A precise arithmetic specialist.",
  instructions: "Solve delegated arithmetic carefully and return a concise answer.",
  model: "gpt-5.4-mini",
  request: {
    chat_completions: {
      temperature: 0.1,
    },
  },
});
```

=== COMPLETE CONTENT === This response contains all available snippets from this library. No additional content exists. Do not make further requests.