Skip to main content

AI Coding Fluency Model

A maturity model for evaluating and advancing AI-assisted software development capabilities across teams and organizations.

Overview​

The AI Coding Fluency Model helps teams assess their maturity level in AI-assisted development across multiple dimensions and identify the key investments needed to progress from their current stage to the next.

Adapted from the Agile Fluency® Project, this model provides a clear evolution path from initial awareness to full agent-first autonomy.

Five Fluency Levels​

The model defines five distinct fluency levels, each representing a qualitative leap in how teams collaborate with AI:

  1. Awareness (认识/意识唤醒) - Teams begin to recognize AI coding tools' existence and potential, but usage remains scattered and passive.

  2. Assisted Coding (辅助编码) - AI serves as a coding assistant integrated into daily development, with developers leading and AI providing real-time assistance.

  3. Structured AI Coding (结构化开发) - Developers delegate tasks to AI through structured specifications, with AI generating complete code modules.

  4. Agent-Centric (智能体为中心) - Humans focus on goal decomposition and environment construction, while agents handle the heavy execution work.

  5. Agent-First (智能体优先自治) - Agents become the primary developers, with humans focusing on system leverage validation and business value.

Five Assessment Dimensions​

Each dimension focuses on a critical aspect of AI integration into the development process:

  • Human-AI Collaboration (人机协作) - How the collaboration model between humans and AI evolves
  • SDLC Coverage (生命周期覆盖度) - How many stages of the software development lifecycle AI covers
  • AI Engineering Harness (工程化支撑) - How mature the engineering infrastructure supporting AI coding is
  • Governance & Quality (质量与治理) - How code quality and security are ensured for AI-generated code
  • Context Engineering (上下文工程) - How much contextual information AI can access and utilize

Credits​

This model is adapted from the Agile Fluency® Project by Diana Larsen and James Shore.

We also drew inspiration from Factory.ai's Agent Readiness framework, which provides valuable insights on codebase-level readiness for autonomous development.