stably plan analyzes your repository and generates user-facing test.fixme() plan files for missing or weak coverage areas.
Use it when you want a concrete testing backlog before generating real tests or filling in gaps manually.
Command Syntax
stably plan accepts an optional prompt. If you omit it, Stably uses a default planning prompt that analyzes the repository and generates a comprehensive test coverage plan.
What It Does
- Reads your codebase and existing tests
- Identifies likely coverage gaps, missing scenarios, and edge cases
- Generates plan files using
test.fixme()so the plan is visible in your repo - Lets you review and refine the plan before turning it into finished tests
- Stays focused on repository analysis rather than browser-driven verification
When To Use It
Use stably plan when you want to… | Why it fits |
|---|---|
| Review coverage before a release | It gives you a structured backlog instead of ad hoc ideas |
| Audit a large repo or legacy test suite | It helps surface missing scenarios without requiring a full rewrite |
| Scope work for a team or agent | test.fixme() plan files are easy to review and split up |
| Plan first, implement later | It stays in planning mode instead of trying to finish tests immediately |
Usage
How It Fits In The Workflow
stably plan is the planning step. stably create is the implementation step.
| Command | Best for |
|---|---|
stably plan | Discovering what should be tested next |
stably create | Generating real Playwright tests |
stably verify | Checking that behavior works in a live browser without creating test files |
stably test | Running your existing suite |
stably fix | Repairing failing tests |
Typical Workflow
test.fixme() files, decide which scenarios matter most, and then use those scenarios to guide stably create or manual test authoring.
Notes
stably planis a repo-analysis command. It does not launch a browser.- It is intended to produce planning artifacts, not complete end-to-end tests.
- The output is meant to be reviewed by a developer or coding agent before being turned into finished tests.
- The exact files generated depend on what the agent finds in your repository, so review the new
test.fixme()plan files after the command completes.
CLI Commands
See the full CLI command overview