| name | big-d:plan |
| description | Create big-d plans that preserve Promise Theory, intent-space purity, and honest promise lifecycles without forking upstream ce:plan |
big-d:plan
Use this for planning work in big-d that touches:
intent-space/
tcp-reference-station/
loop/
- promise lifecycles
- agent control surfaces
- managed coordination or onboarding flows
Why this exists
Upstream ce:plan remains the canonical planner.
This wrapper is intentionally thin. Its job is to delegate to the normal ce:plan workflow first, then augment the resulting plan with the local big-d guardrails so plans do not drift away from:
- autonomous participants
- promises about self
- honest promise lifecycles
- intent-space purity
- explicit assessment boundaries
Read first
docs/architecture/promise-native-planning-guardrails.md
docs/checklists/promise-native-plan-review.md
- Relevant subproject docs from
AGENTS.md
Workflow
- First execute the normal
ce:plan workflow for the request.
- After that draft exists, apply the local
big-d guardrails before finalizing it.
- Add or refine a section named
## Promise-Native Architecture Check.
- In that section, answer:
- who the autonomous participants are
- which promises are about self
- where state authority lives
- which lifecycle acts are required and why
- how the design preserves intent-space purity
- how visibility / containment is scoped
- which shortcut was rejected because it violated the stance
- Run the checklist in
docs/checklists/promise-native-plan-review.md.
- If the checklist fails, revise the plan before implementation.
Keep the wrapper policy-focused. Do not replace ce:plan with a separate custom planning process unless upstream ce:plan is unavailable.
Fallback if this skill is not auto-discovered
If your tool does not auto-load repo-local skills, read this file directly and follow it manually. The source of truth is the repo, not a machine-local skill install.