| name | spae-build |
| description | Atomic Execution for the `SPAE` framework. Executes exactly one atomic task from `PLAN.md`. |
| user-invocable | true |
| argument-hint | [optional-workstream-name] e.g. 'user-auth' |
Build (SPAE)
Goal: execute exactly one atomic task from PLAN.md.
When to use
- When
STATE.json reaches phase: build.
- To execute the next task in the plan.
Process
- Initialize: Resolve the
workstream via the provided name or the
.spae/current symlink. Read STATE.json and PLAN.md to identify
the active task.
- Execute:
- Write the minimal code changes required to meet the task's
acceptance criteria.
- Restrict edits to relevant files; avoid incidental refactoring or
cleanup.
- Invoke
vibe_check with goal (current task name) and plan
(numbered implementation steps) for high-complexity or
system-altering tasks.
- Add required tests and run the verification steps defined in
PLAN.md.
- Finalize:
- Mark the task
done and increment completion metrics in
STATE.json.
- Advance the cursor to the next task or set
phase: verify if the
plan concludes.
- Output the standardized Execution Summary and Task Execution
Feedback (or Phase Transition Feedback if the plan
concludes).
Verification
- Task acceptance criteria met.
- Verification steps pass.
STATE.json updated correctly.
Rules
- Optimize all operations for agent, token, and context efficiency.
- Execute the active task with maximal signal and minimal edits.
- Execute exactly one atomic task per invocation.
- Write Boundaries: Exercise exclusive authority to edit source code
and non-
SPAE project files.
- Forbidden Writes: Never edit
PLAN.md or SPEC.md during this
phase.
- If a task fails, report the blocker and halt.
- STATUS: SUCCESS on completion.
Testing rules
- Adopt a test-first, verify-last approach.
- Write exhaustive unit tests covering edge cases.
- Write acceptance and integration tests where relevant to the task.
- Ensure all test suites pass before finalizing the task.
Standardized feedback
- Keep feedback prose terse, concise, and precise.
- Optimize prose for token and context efficiency.
- If needed, split findings and summary into terse bullet points.
### Execution Summary
- **Actions**:
- [List of terse, short, compact, condensed summary of actions taken]
- **Files**:
- [List of modified or created files]
- **Findings**:
- [List of terse summary of key gaps, risks, or architectural notes]
- **Summary**:
- [List of terse summary of changes]
Task execution feedback
> **SPAE Status** • `workstream-name`
> **Progress**: Task [X] of [Y] ([Z] remaining)
> **Completed**: `T-XXX` - [Task title]
> **Next Task**: `T-YYY` - [Next task title]
>
> _Run `/build` (or `/tdd`) to execute the next task._
Phase transition feedback (on plan conclusion)
> **SPAE Status** • `workstream-name`
> **Progress**: All [X] tasks completed
> **Phase Complete**: `/build`
> **Next Phase**: `/verify`
>
> _Run `/verify` next._