Adoption

How to adopt this standard in a new or existing repository. Three paths, one per profile.

See profiles.md for the exact file lists and standard.md for the conceptual definition.

Adopt OSS-only profile

Use this when you want a credible open-source baseline without AI-specific files.

  1. Start from the template. Click “Use this template” on the repository’s GitHub page (available from v0.2.0).
  2. Remove AI-specific paths. Delete AGENTS.md, CLAUDE.md, .github/copilot-instructions.md, .cursor/, .ai/, docs/ai-workspace-layer.md, docs/practical-ai-collaboration.md, docs/ai-tools.md.
  3. Adjust the README. Remove the AI-workspace and AI-collaboration sections.
  4. Replace project metadata. Update name, description, license holder, security contact, roadmap, and changelog.
  5. Configure CI. Keep .github/workflows/ci.yml; the structural checks still apply.
  6. Apply branch protection. Follow docs/branch-protection.md.

Adopt AI-only profile

Use this when you already have an open-source or internal repository and want to introduce structured Human-AI Collaboration without re-doing the OSS scaffolding.

  1. Copy the AI files into your existing repository. Copy AGENTS.md, CLAUDE.md, .github/copilot-instructions.md, .cursor/rules/00-project.mdc, the full .ai/ directory, and the AI docs (docs/ai-workspace-layer.md, docs/practical-ai-collaboration.md, docs/ai-tools.md).
  2. Localize AGENTS.md. Adjust the approval rules, branch-naming conventions, and any references to match your existing workflow.
  3. Seed .ai/. Replace placeholder content in .ai/project-brief.md, .ai/state.md, and .ai/decisions.md with your project’s actual context.
  4. Add a short pointer in your existing README.md. Mention AGENTS.md and .ai/ so contributors and AI tools discover them.
  5. Optional: pull in docs/standard.md, docs/profiles.md, and docs/adoption.md if you want to expose the standard’s structure to your team.

Adopt combined profile

Use this when you are starting a new project and want both baselines from day one.

  1. Use the GitHub template. Click “Use this template” to create your repository from this one.
  2. Replace project metadata. Name, description, license holder, security contact, roadmap.
  3. Seed .ai/. Update project-brief.md, state.md, and decisions.md with your project’s reality.
  4. Apply branch protection per docs/branch-protection.md.
  5. Decide on AI tools. Keep the pointer files for the tools you use; delete the others (or keep them as future-ready stubs — they are small).
  6. Iterate. Open issues and PRs as normal; the standard is a starting point, not a constraint.

Updating from a previous version

The standard follows Semantic Versioning (see CONTRIBUTING.md → Versioning).

FromToAction
v0.1.0v0.2.0Pull in new files: docs/standard.md, docs/profiles.md, docs/adoption.md, docs/ai-tools.md, .cursor/rules/00-project.mdc. Update README.md and AGENTS.md per upstream diff. No path renames.

Recommended workflow for adopters tracking the standard:

  1. Add this repository as a remote: git remote add upstream https://github.com/rubennati/ai-project-standard.git.
  2. Fetch a release tag: git fetch upstream --tags.
  3. Diff the desired version against your current adoption: git diff <your-base>..upstream/v0.2.0 -- AGENTS.md docs/ .ai/.
  4. Cherry-pick or copy the changes that apply to your project. Skip the rest.

Breaking changes for adopters are always called out in the corresponding CHANGELOG.md entry under ### Breaking changes for adopters.


Source: docs/adoption.md — edits land here on the next deploy.