Kimi CLI Tutorial: Multi-Mode Terminal Agent with MCP and ACP

May 11, 2026 ยท View on GitHub

Learn how to use MoonshotAI/kimi-cli to run an interactive terminal coding agent with configurable modes, MCP integrations, and ACP-based IDE connectivity.

GitHub Repo License Release

Why This Track Matters

Kimi CLI combines strong terminal workflows with MCP, ACP, session controls, and automation-oriented modes, making it a high-impact option for agentic development teams.

This track focuses on:

  • terminal setup, providers, and first-run flow
  • command and mode controls for interactive/non-interactive use
  • skills, agents, and MCP extension paths
  • IDE connectivity through ACP and production governance

Current Snapshot (auto-updated)

Mental Model

flowchart LR
    A[Prompt or Command] --> B[Kimi CLI Core]
    B --> C[Agent + Tools]
    C --> D[MCP / External Integrations]
    C --> E[Session and Context Controls]
    B --> F[ACP / IDE Clients]

Chapter Guide

ChapterKey QuestionOutcome
01 - Getting StartedHow do I install and run Kimi CLI quickly?Working baseline
02 - Command Surface and Session ControlsWhich startup flags and session options matter most?Strong command mental model
03 - Agents, Subagents, and SkillsHow do I customize behavior for team workflows?Better task specialization
04 - MCP Tooling and Security ModelHow do I add MCP servers safely?Extensibility with controls
05 - ACP and IDE IntegrationsHow does Kimi connect to external clients and IDEs?Multi-client integration strategy
06 - Shell Mode, Print Mode, and Wire ModeWhich mode should I use for interactive work vs automation?Mode-selection framework
07 - Loop Control, Retries, and Long TasksHow do I tune long-running autonomous behavior?Better reliability under complex tasks
08 - Production Operations and GovernanceHow do teams standardize Kimi CLI usage safely?Operational baseline

What You Will Learn

  • how to run Kimi CLI across interactive, print, and protocol modes
  • how to customize agents/skills for project-specific workflows
  • how to integrate MCP tools and ACP clients with proper controls
  • how to operationalize Kimi CLI for team-scale development

Source References


Start with Chapter 1: Getting Started.

Full Chapter Map

  1. Chapter 1: Getting Started
  2. Chapter 2: Command Surface and Session Controls
  3. Chapter 3: Agents, Subagents, and Skills
  4. Chapter 4: MCP Tooling and Security Model
  5. Chapter 5: ACP and IDE Integrations
  6. Chapter 6: Shell Mode, Print Mode, and Wire Mode
  7. Chapter 7: Loop Control, Retries, and Long Tasks
  8. Chapter 8: Production Operations and Governance

Generated by AI Codebase Knowledge Builder