STARTER REPO TO MAKE AI CODING TOOLS FOLLOW YOUR CI AND TESTS
An open-source starter repo ties Python linting, tests, and AI-assistant rules together so code from tools like Cursor, Claude Code, Codex, and GitHub Copilot a...
An open-source starter repo ties Python linting, tests, and AI-assistant rules together so code from tools like Cursor, Claude Code, Codex, and GitHub Copilot aligns with your team's standards.
A detailed write-up shows how instruction files (e.g., AGENTS.md, CLAUDE.md, and .cursor/rules/) guide AI assistants to follow the same conventions enforced in CI, turning them from context-free generators into project-aware contributors; see the overview on DEV.
The repo’s GitHub Actions run Ruff with --fix (auto-committing changes), then pytest, and post PR comments that name the exact test files, classes, and method signatures required for changed sources, with pre-commit hooks and incident runbooks rounding out day-1 guardrails; grab the code at github.com/humzakt/dev-starter-kit.
Codifies AI-assisted coding into your SDLC so generated code adheres to style, testing, and workflow rules.
Reduces review toil by auto-fixing lint, enforcing test coverage, and giving precise PR guidance.
-
terminal
Verify each AI assistant respects AGENTS.md and tool-specific rules by generating changes and checking CI outcomes.
-
terminal
Confirm CI auto-fix commits are idempotent and that PR test prompts correctly map to changed source files.
Legacy codebase integration strategies...
- 01.
Adopt incrementally by mapping Ruff/pytest settings to your existing linters/tests and merging the PR Checks workflow.
- 02.
Watch for conflicts with current pre-commit and GitHub Actions; stage changes behind branch protection on a pilot repo.
Fresh architecture paradigms...
- 01.
Template new services from the repo to lock in conventions once and propagate them with AI tool configs and CI from day one.
- 02.
Use the AI instruction files as part of onboarding so assistants default to your architecture and runbooks.