# AI development environment decision matrix

Score each route: 1 = blocking risk, 2 = acceptable, 3 = good fit. Add links to official documentation.

| Criterion | Importance 1–3 | Visual builder | Coding agent in repo | IDE + AI assistant | Evidence / source |
|---|---:|---:|---:|---:|---|
| Supports my user journey | | | | | |
| Source can be saved/moved | | | | | |
| Changed files and diff are visible | | | | | |
| Errors and tests can be checked | | | | | |
| Data location is clear | | | | | |
| Access model is clear | | | | | |
| Previous version can be restored | | | | | |
| Another specialist can continue | | | | | |

**Choice:**
**Why:**
**Known limits:**
**Small test before moving the real project:**
**How to undo a bad change:**