From rtl-agent-team
Orchestrates Phase 2 RTL architecture design using Claude Code native teams for parallel HW candidate evaluation, dual-stream architecture, RefC development, and 3-round iterative reviews.
npx claudepluginhub babyworm/rtl-agent-team --plugin rtl-agent-teamThis skill is limited to using the following tools:
<Purpose>
Generates design tokens/docs from CSS/Tailwind/styled-components codebases, audits visual consistency across 10 dimensions, detects AI slop in UI.
Records polished WebM UI demo videos of web apps using Playwright with cursor overlay, natural pacing, and three-phase scripting. Activates for demo, walkthrough, screen recording, or tutorial requests.
Delivers idiomatic Kotlin patterns for null safety, immutability, sealed classes, coroutines, Flows, extensions, DSL builders, and Gradle DSL. Use when writing, reviewing, refactoring, or designing Kotlin code.
<Use_When>
<Do_Not_Use_When>
Phase 1 completion required:
docs/phase-1-research/iron-requirements.json must existdocs/phase-1-research/open-requirements.json (optional — absent if P1 had no open items)docs/phase-1-research/io_definition.json must existIf prerequisites are missing: WARNING — recommend running /rtl-agent-team:p1-spec-research.
Proceed with available artifacts — orchestrator will adapt scope.
# Step 1: Team creation (main session = leader)
TeamCreate(team_name="p2-arch", description="Phase 2 architecture: dual-stream arch + RefC development")
# Step 2: Write team-config.json for hook consumption
Write(".rat/state/team-config.json", json.dumps({
"team_mode": true,
"team_name": "p2-arch",
"leader_session_id": "<current_session_id>",
"coordinator_name": "coordinator",
"worker_count": 3,
"phase": "p2",
"created_at": "<ISO_TIMESTAMP>"
}))
# Step 3: Prepare directories
Bash("mkdir -p docs/phase-2-architecture reviews/phase-2-architecture .rat/scratch/phase-2")
# Step 4: No initial tasks from skill — coordinator creates T1a-N after reading P1 artifacts
# Step 5: Spawn coordinator as teammate (orchestrator)
Agent(team_name="p2-arch", subagent_type="rtl-agent-team:p2-arch-team-orchestrator",
name="coordinator", description="P2 architecture coordination",
prompt="You are the Phase 2 architecture coordinator in team 'p2-arch'. "
"Manage the task graph using TaskCreate/TaskList/TaskUpdate. "
"Direct workers via SendMessage. "
"Create all tasks (T1a-N HW eval through T13 final consolidation). "
"Signal leader when T13 complete. User input: $ARGUMENTS")
# Step 6: Spawn workers as teammates (3 general-purpose)
Agent(team_name="p2-arch", subagent_type="rtl-agent-team:arch-designer",
name="arch-worker", description="P2 architecture design and review",
prompt="You are a Phase 2 architecture worker in team 'p2-arch'. "
"Coordinator: 'coordinator' (send results via SendMessage). "
"Phase artifacts: docs/phase-2-architecture/, reviews/phase-2-architecture/. "
"Specialty: architecture design, HW evaluation, bandwidth integration. "
"For specialist work, spawn: Task(subagent_type='rtl-agent-team:<specialist>', prompt='...'). "
"Examples: vcodec-architecture-expert for HW eval, rtl-architect for review. "
"Follow Team Worker Protocol in agents/lib/team-worker-preamble.md. "
"Scratch dir: .rat/scratch/phase-2/ (for write-restricted outputs).")
Agent(team_name="p2-arch", subagent_type="rtl-agent-team:ref-model-dev",
name="refmodel-worker", description="P2 reference model development",
prompt="You are a Phase 2 ref-model worker in team 'p2-arch'. "
"Coordinator: 'coordinator' (send results via SendMessage). "
"Phase artifacts: refc/, docs/phase-2-architecture/. "
"Specialty: C reference model development, model consistency review. "
"For specialist work, spawn: Task(subagent_type='rtl-agent-team:ref-model-reviewer', prompt='...'). "
"Follow Team Worker Protocol in agents/lib/team-worker-preamble.md.")
Agent(team_name="p2-arch", subagent_type="rtl-agent-team:rtl-architect",
name="review-worker", description="P2 architecture review lead",
prompt="You are a Phase 2 review worker in team 'p2-arch'. "
"Coordinator: 'coordinator' (send results via SendMessage). "
"Phase artifacts: reviews/phase-2-architecture/. "
"Specialty: spec compliance review, architecture review aggregation. "
"For specialist work, spawn: Task(subagent_type='rtl-agent-team:<specialist>', prompt='...'). "
"Follow Team Worker Protocol in agents/lib/team-worker-preamble.md. "
"Scratch dir: .rat/scratch/phase-2/ (for write-restricted outputs).")
# Step 7: Leader monitoring loop — poll until all tasks complete
while True:
tasks = TaskList()
all_done = all(t.status == "completed" for t in tasks)
if all_done:
break
# Continue polling
# Step 8: Cleanup
TeamDelete()
Bash("rm -f .rat/state/team-config.json")
Bash("rm -rf .rat/scratch/phase-2/")
open-requirements.jsonopen-resolved and ambiguity-passcompliance-pass will be added when team orchestrator is updated to invoke compliance-checker