---
url: /en/ai/intro.md
description: >
  Introduces the AI features available on the CNB platform, including CodeBuddy
  NPC, TencentCloud CodeBuddy integration, Chat assistant, RAG Knowledge Base,
  Issue Analysis, and Change Analysis.
---
The CNB platform includes several built-in AI features that can boost developer productivity.

## Features

### CodeBuddy NPC

[CodeBuddy NPC](../build/npc.md) is the AI colleague living in your CNB dev infrastructure.

Set a goal for the NPC, review its output — a Cloud Agent built on CodeBuddy,
deeply integrated with the CNB platform, autonomously acquiring context
and completing dev tasks end-to-end.

**Core Capabilities**:

* **Goal-Driven**: You define "what", the NPC handles "how".
  From requirements to PR, fully autonomous.
* **Autonomous Context**: Deeply integrated with CNB repos, issues, and pipelines.
  Understands requirements, reads code, checks logs — no manual intervention needed.
* **Parallel Assignment**: Dispatch multiple tasks simultaneously.
  Multiple NPCs work in parallel in the cloud, unblocked and independently delivered.

**Quick Start**:

* Mention `@CodeBuddy` in an Issue or PR comment, and the NPC automatically
  executes the task and replies
* Define a role in your repository's `.cnb/settings.yml` to create your own NPC,
  then use it via `@repo-path(role-name)`
* After enabling [Work Mode](../build/npc.md#work-mode), the NPC can autonomously
  write code, push code, and create PRs

For more information on event triggers and custom NPCs,
see [NPC](../build/npc.md).

### TencentCloud CodeBuddy

[TencentCloud CodeBuddy](https://codebuddy.cn/) is built into the Workspaces WebIDE,
ready to use without additional installation.

### AI Chat

The platform has a built-in
[AI Chat](//cnbprod.alpha.qq.com/?aiq=CNB%20AI%20Function%20Introduction),
supporting quick access to help documentation and configuration file writing through AI.

### RAG Knowledge Base

Set up enterprise or personal knowledge bases through code repositories.
After committing documents to the repository and configuring the pipeline,
document content is processed by the large model and
automatically written to the knowledge base for webpage Q\&A and OpenAPI scenarios,
helping build RAG (Retrieval-Augmented Generation) applications.
See [Knowledge Base](./knowledge-base.md) for details.

### Issue Analysis

Submit build error messages to AI with one click to quickly identify the root cause.

### Change Analysis

Submit commit or pull request change information to AI with one click
to quickly understand changes and their potential impact.