How this skill is triggered — by the user, by Claude, or both
Slash command
/pasture:architect-propose-planThe summary Claude sees in its skill listing — used to decide when to auto-load this skill
<!-- BEGIN GENERATED FROM pasture schema -->
Command: pasture:architect:propose-plan — Create PROPOSAL-N task with full technical plan
-> Full workflow in PROCESS.md <- Phase 3
[arch-propose-bdd-format]
[arch-propose-checklist-required]
[arch-propose-revision-history]
pasture:supersededStarting new feature design; creating formal plan for review.
Proposals are numbered incrementally: PROPOSAL-1, PROPOSAL-2, etc. Each revision increments N. Old proposals are marked pasture:superseded with a comment explaining why.
bd create --type=feature \
--labels="pasture:p3-plan:s3-propose" \
--title="PROPOSAL-1: <feature name>" \
--description="$(cat <<'EOF'
---
references:
request: <request-id>
urd: <urd-id>
---
## Problem Space
**Axes of the problem:**
- Parallelism: ...
- Distribution: ...
**Has-a / Is-a:**
- X HAS-A Y
- Z IS-A W
## Engineering Tradeoffs
| Option | Pros | Cons | Decision |
|--------|------|------|----------|
| A | ... | ... | Selected |
| B | ... | ... | Rejected |
## MVP Milestone
<scope with tradeoff rationale>
## Public Interfaces
\`\`\`go
type Example interface { /* ... */ }
\`\`\`
## Types & Enums
\`\`\`go
type ExampleType int
const (
ExampleTypeA ExampleType = iota
ExampleTypeB
)
\`\`\`
## Validation Checklist
### Phase 1
- [ ] Item 1
- [ ] Item 2
### Phase 2
- [ ] Item 3
## BDD Acceptance Criteria
**Given** precondition
**When** action
**Then** outcome
**Should Not** negative case
## Files Affected
- pkg/path/file1.go (create)
- pkg/path/file2.go (modify)
EOF
)" \
--design='{"validation_checklist":["Item 1","Item 2","Item 3"],"tradeoffs":[{"decision":"Use A","rationale":"Because..."}],"acceptance_criteria":[{"given":"X","when":"Y","then":"Z","should_not":"W"}]}'
# Link to request
bd dep add <request-id> --blocked-by <proposal-id>
Read the URD and Phase 1 outputs to understand full context before drafting:
bd show <urd-id>
bd show <request-id> # includes classification, research findings, explore findings as comments
The URD contains the structured requirements, priorities, design choices, and MVP goals from the URE survey. The REQUEST task comments contain Phase 1 outputs: classification (4 axes), domain research findings (prior art, standards), and codebase exploration findings (entry points, related types, dependencies). Your proposal must:
<urd-id>After creating PROPOSAL-N task:
/pasture:architect-request-review to spawn 3 reviewers/pasture:architect-ratify to add ratify label to PROPOSAL-NWhen creating proposals for a follow-up epic (received via h6 from supervisor):
FOLLOWUP_PROPOSAL-N: (e.g., FOLLOWUP_PROPOSAL-1: Add request-id correlation)original_urd: <id> and followup_urd: <id> in frontmatternpx claudepluginhub dayvidpham/aura-plugins --plugin pastureGuides creation, editing, and verification of skills for AI coding agents using test-driven development with subagent scenarios. Use when authoring or debugging skills.