| name | spec-for-openclaw |
| description | Define implementation scope before coding inside an OpenClaw workspace. Use when the request sounds like "write a spec", "define this feature first", "clarify scope", "turn this idea into something buildable", or any task where outcome, constraints, interfaces, or acceptance criteria are still fuzzy. |
Spec For OpenClaw
Write the implementation target down before code starts.
Process
- State the problem in concrete terms.
- Define the desired outcome from the user's perspective.
- List constraints from the workspace, runtime, tools, and deployment surface.
- Identify affected files, systems, and interfaces.
- Write acceptance criteria that can later be verified.
- Call out unknowns that must be resolved before broad implementation.
Output Standard
A useful spec is short, testable, and specific enough that planning becomes straightforward.
Include:
- objective
- scope
- non-goals
- affected surfaces
- acceptance criteria
- verification idea
Anti-Rationalization
| Rationalization | Response |
|---|
| "The task is obvious." | If it is obvious, it should be easy to state clearly. |
| "I can spec while coding." | That usually hides scope drift. Define the target first. |
| "This is just a bug fix." | Bugs still need expected behavior and proof. |