# Adopting Assistants in a Team — AI Coding Assistants: Use Them Well and Review Them Hard

Source: https://www.geekswithgeeks.com/en/ai-coding-assistants/learn-team-adoption

> Share conventions, keep review standards and measure real outcomes.

## Rules of the road

Teams do best when they agree on a few things: which **tools and data policies** are approved, a shared **project instruction file** (for example `AGENTS.md` or `CLAUDE.md`) listing build and test commands and coding conventions, the rule that **the author owns the code** regardless of who typed it, and that AI-assisted pull requests get the same review as any other. Measure real outcomes such as review time, defects and cycle time, not only lines produced, because more code is not the same as more value.

## A project instruction file

A short file read by assistants at the start of a session. Keep it accurate and under review like any other code.

```markdown
# Project guide for AI assistants

## Commands
- Install: `npm ci`
- Test: `npm test`   Lint: `npm run lint`

## Conventions
- TypeScript strict mode; no `any`.
- New endpoints need a test in `tests/api/`.
- Never edit files under `src/generated/`.
- Never commit secrets; use `.env` (git-ignored).
```

**Quiz:** Who is responsible for AI-generated code in a pull request?

- [x] The author who submits it
- [ ] The AI tool vendor
- [ ] Nobody
- [ ] The reviewer only

*Answer:* The author who submits it. Whoever commits code owns it, however it was produced.