[Initiative]: State of AI-Assisted Development in CNCF Projects

[graz-dev]

Name

State of AI-Assisted Development in CNCF Projects

Short description

This initiative empowers the CNCF projects to move from ad-hoc experimentation to confident adoption by surveying real-world AI tool usage across CNCF projects and delivering a practical, vendor-neutral guide for secure integration.

Responsible group

TAG Developer Experience

Does the initiative belong to a subproject?

No

Subproject name

No response

Primary contact

Graziano Casto (gh: @graz-dev | email: cst.grzn@gmail.com)

Additional contacts

No response

Initiative description

This initiative establishes a comprehensive, observational study of AI-assisted development practices across the entire CNCF ecosystem, targeting projects from Sandbox to Graduated maturity levels. The scope is strictly defined to analyze real-world usage within the software development lifecycle spanning local IDE environments, CI pipelines, and code review processes and more while explicitly excluding any engineering efforts to build tools, train models, or conduct commercial performance benchmarking. Spanning a structured six-month timeline, the initiative will drive the transition from ad-hoc experimentation to confident adoption through the following specific deliverables:

• Month 1: The "State of AI-Assisted Development" Survey: Design and launch a wide-reaching survey aimed at securing responses from at least 30% of CNCF projects. This deliverable focuses on capturing raw data regarding tool selection, usage frequency, and integration points across the community.

• Month 3: Data-Driven Analysis Report: Publication of a public report that synthesizes the survey findings. This document will identify the top 5–10 tools currently in use, map common workflow patterns, and provide a measurable assessment of perceived benefits and security challenges.

• Month 6: Vendor-Neutral Reference Guide The final deliverable is an actionable handbook for maintainers. Based on the analysis, this guide will provide concrete recommendations on security, productivity, and edge-case management, offering a standard set of best practices for integrating AI tools into cloud native project workflows.

By bridging the gap between isolated experimentation and collective intelligence, this initiative delivers immediate, high-impact value to the CNCF ecosystem. It replaces the current uncertainty of AI adoption with a trusted, data-backed roadmap, empowering maintainers to integrate new tools with confidence regarding security and workflow efficiency. Ultimately, this work elevates the standard of the developer experience across all projects, transforming AI-assisted development from a fragmented challenge into a streamlined, secure, and powerful asset for the entire community.

Deliverable(s) or exit criteria

• The "State of AI-Assisted Development" Survey: Design and launch a wide-reaching survey aimed at securing responses from at least 30% of CNCF projects. This deliverable focuses on capturing raw data regarding tool selection, usage frequency, and integration points across the community.
• Data-Driven Analysis Report: Publication of a public report that synthesizes the survey findings. This document will identify the top 5–10 tools currently in use, map common workflow patterns, and provide a measurable assessment of perceived benefits and security challenges.
• Vendor-Neutral Reference Guide The final deliverable is an actionable handbook for maintainers. Based on the analysis, this guide will provide concrete recommendations on security, productivity, and edge-case management, offering a standard set of best practices for integrating AI tools into cloud native project workflows.

Tracking document for meeting and progress

https://notes.cncf.io/s/wCOGYaJcw#

[pavanipriyasajja]

I am interested to contribute as a UX designer, I will also work on the user research for the project to understand the actual user problems and create a developer friendly solutions.

[ooye-sanket]

Hi @graz-dev 👋
I would love to contribute to this initiative.

I’m a CNCF community core member ( CNCF Ahmedabad) and also work on with cloud-native development, AI-assisted tooling, and developer workflows across multiple open-source projects.

I can contribute in the following areas:

• Survey design & distribution across CNCF project communities
• Data collection and analysis
• Drafting the final vendor-neutral reference guide
• Documenting AI workflow patterns used in open-source projects

I’m familiar with AI-assisted coding tools, GitHub workflows, CI/CD automation, and dev productivity metrics, so I can support the initiative from both a DX and technical perspective.

Let me know how I can get involved in the working group and contribute from the next steps.

THANKS!

[joshuabezaleel]

Eager to see how this will proceeds! 🎉 Some initial thoughts:

1. As open source is the embodiment of collective software work, and AI agentic coding has taken the field by the storm, I think it will be inevitable that each corresponding contributors/maintainers of each projects would be thinking to make an attempt for it, so this might be beneficiail as a governance/guidance/starter kit for projects.
2. Might need to work as well with Contributor Strategy Subproject/Maintainers Circle, not sure how the process is after the reboot, and how they coordinate with each of the projects.
3. One example of the point 1 earlier was that we can see AGENTS.md file in various Apache projects (https://github.com/apache/incubator-hugegraph/blob/master/AGENTS.md, https://github.com/apache/gravitino/blob/main/AGENTS.md, https://github.com/apache/airflow/blob/main/AGENTS.md) which means that the contributors use this for AI coding. But for now I can't seem to find yet whether Apache foundation has any specific guidance/governance on this kind of thing, or is it free for each projects to determine themselves.