Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx skills add babyworm/rtl-agent-team --skill rtl-p4-implement-teamgit clone --depth 1 https://github.com/babyworm/rtl-agent-teamWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/skills/babyworm/rtl-agent-team/rtl-p4-implement-team)<a href="https://agentmods.dev/skills/babyworm/rtl-agent-team/rtl-p4-implement-team"><img src="https://agentmods.dev/badge/skills/babyworm/rtl-agent-team/rtl-p4-implement-team/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/skills/babyworm/rtl-agent-team/rtl-p4-implement-team"><img src="https://agentmods.dev/badge/skills/babyworm/rtl-agent-team/rtl-p4-implement-team.svg" alt="Reviewed on agentmods" width="80" height="20"></a>What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00039 | $0.01778 |
| Opus 5.5 | $0.00016 | $0.00711 |
| Sonnet 5.5 | $0.00008 | $0.00356 |
| Haiku 4.5 | $0.00004 | $0.00178 |
Grade C, and why
rtl-p4-implement-team scanned grade C with 1 finding against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 21d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Recursive force deletehighDestructive command
rm -rf with a variable or a broad path is one typo away from removing the wrong tree.
Bash("rm -rf .rat/scratch/phase-4/") How it starts
The opening of the file, as written. The whole thing — 136 lines — stays where its author put it; the contents beside it link to each section on GitHub.
<Use_When>
- Phase 3 μArch specs are complete and RTL coding is needed
- User says "implement team", "Phase 4 team", "parallel implement"
- Have 3+ modules that benefit from concurrent RTL coding workers
- Need maximum parallelism for multi-module RTL implementation </Use_When>
<Do_Not_Use_When>
- Phase 3 μArch specs are not complete (run rtl-p3-uarch-design first)
- Single module only (use rtl-p4-implement for simpler flow)
- Only need a single module bug fix (use rtl-p4s-bugfix)
- Need RTL + verification together from uarch (use rat-p4p5-impl-verify) </Do_Not_Use_When>
Prerequisites
Phase 3 completion required:
- At least one μArch spec in
docs/phase-3-uarch/must exist
If prerequisite is missing: WARNING — recommend running /rtl-agent-team:rtl-p3-uarch-design.
Proceed with available artifacts — orchestrator will adapt scope.
Execution
# Step 1: Team creation (main session = leader)
TeamCreate(team_name="p4-implement", description="Phase 4 RTL implementation: 10-wave parallel module coding")
# Step 2: Write team-config.json for hook consumption
Write(".rat/state/team-config.json", json.dumps({
"team_mode": true,
"team_name": "p4-implement",
"leader_session_id": "<current_session_id>",
"coordinator_name": "coordinator",
"worker_count": 4,
"phase": "p4",
"created_at": "<ISO_TIMESTAMP>"
}))
# Step 3: Prepare directories
Bash("mkdir -p reviews/phase-4-rtl docs/phase-4-rtl .rat/scratch/phase-4")
# Step 4: No initial tasks from skill — coordinator creates per-module W1-W10 after reading uarch specs
# Step 5: Spawn coordinator as teammate (orchestrator)
Agent(team_name="p4-implement", subagent_type="rtl-agent-team:p4-implement-team-orchestrator",
name="coordinator", description="P4 implementation coordination",
prompt="You are the Phase 4 RTL implementation coordinator in team 'p4-implement'. "
"Manage the 10-wave task graph using TaskCreate/TaskList/TaskUpdate. "
"Direct workers via SendMessage. "
"Create per-module 10-wave task graph after reading uarch specs. "
"Signal leader ONLY after the Phase 4 Gate (Step 4 — all required checks "
"and artifacts, not merely the W10 integration gate) passes, including the "
"mandatory Codex cross-review. User input: $ARGUMENTS")
# Step 6: Spawn workers as teammates (4 general-purpose)
Agent(team_name="p4-implement", subagent_type="rtl-agent-team:rtl-coder",
name="coder-0", description="P4 RTL coding and lint",
prompt="You are a Phase 4 coding worker in team 'p4-implement'. "
"Coordinator: 'coordinator' (send results via SendMessage). "
"Phase artifacts: rtl/{module}/{module}.sv, reviews/phase-4-rtl/. "
"Specialty: RTL module implementation (W1), lint (W2), lint fix (W3). "
"For specialist work, spawn: Task(subagent_type='rtl-agent-team:<specialist>', prompt='...'). "
"Examples: lint-checker for lint, eda-runner for simulation. "
"Follow the Team Worker Protocol section of your agent definition. "
"Naming: i_/o_ prefixes, snake_case, clk/{domain}_clk, rst_n/{domain}_rst_n.")
Agent(team_name="p4-implement", subagent_type="rtl-agent-team:rtl-coder",
name="coder-1", description="P4 RTL coding and lint",
prompt="You are a Phase 4 coding worker in team 'p4-implement'. "
"Coordinator: 'coordinator' (send results via SendMessage). "
"Phase artifacts: rtl/{module}/{module}.sv, reviews/phase-4-rtl/. "
"Specialty: RTL module implementation (W1), lint (W2), lint fix (W3). "
"For specialist work, spawn: Task(subagent_type='rtl-agent-team:<specialist>', prompt='...'). "
"Follow the Team Worker Protocol section of your agent definition. "
"Naming: i_/o_ prefixes, snake_case, clk/{domain}_clk, rst_n/{domain}_rst_n.")
Agent(team_name="p4-implement", subagent_type="rtl-agent-team:testbench-dev",
name="verify-worker", description="P4 unit test and verification",
prompt="You are a Phase 4 verification worker in team 'p4-implement'. "
"Coordinator: 'coordinator' (send results via SendMessage). "
"Phase artifacts: sim/{module}/, reviews/phase-4-rtl/. "
"Specialty: unit test (W6), CDC analysis (W7), protocol check (W8). "
"For specialist work, spawn: Task(subagent_type='rtl-agent-team:<specialist>', prompt='...'). "
"Examples: eda-runner for sim, cdc-checker for CDC, protocol-checker for W8. "
"Follow the Team Worker Protocol section of your agent definition.")
Agent(team_name="p4-implement", subagent_type="rtl-agent-team:rtl-critic",
name="review-worker", description="P4 code review and refactor",
prompt="You are a Phase 4 review worker in team 'p4-implement'. "
"Coordinator: 'coordinator' (send results via SendMessage). "
"Phase artifacts: reviews/phase-4-rtl/. "
"Specialty: code review (W4), bugfix (W5), refactor (W9), integration (W10). "
"For specialist work, spawn: Task(subagent_type='rtl-agent-team:<specialist>', prompt='...'). "
"Follow the Team Worker Protocol section of your agent definition.")
# Step 7: Leader wait — teardown is gated on the COORDINATOR SIGNAL, not on TaskList.
# "All tasks completed" is necessary but NOT sufficient: the coordinator's post-gate
# work (compliance check, ADR generation, phase summary, Codex cross-review) runs via
# direct Task() calls that never appear in the shared task graph.
while True:
if coordinator_signaled_phase_complete:
# SendMessage from 'coordinator' — sent only after the phase gate
# AND all post-gate mandatory steps pass (see coordinator prompt above)
break
tasks = TaskList()
if all(t.status == "completed" for t in tasks):
# Task graph drained but no coordinator signal yet → the coordinator is
# still running gate/post-gate steps. KEEP WAITING — do NOT TeamDelete.
pass
# Continue polling
# Step 8: Cleanup
TeamDelete()
Bash("rm -f .rat/state/team-config.json")
Bash("rm -rf .rat/scratch/phase-4/")
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 21d ago First seen · 136 lines · 39 tokens per session scan C 3ee0b226dd95
rtl-p4-implement-team is a skill published in the GitHub repository babyworm/rtl-agent-team (63 stars, last pushed 21d ago), licensed MIT. It adds 39 tokens to every session and 1,778 once invoked, about $0.0002 per session on Opus 5.5. A static security scan graded it C with 1 finding (recursive force delete). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-10-08.
Other skills, from other repositories
dhx-components-sourcing
Real-time global electronic components sourcing, BOM matching, cross-referencing, and China PCBA factory sourcing agent powered by DHX Tech.
dhx-components-sourcing
Real-time global electronic components sourcing, BOM matching, cross-referencing, and China PCBA factory sourcing agent powered by DHX Tech.
onboarding
Connect a new user's accounts, learn their real work, choose a voice, and set up concrete help.
google-workspace
Read and act on the user's Gmail, Google Calendar, and Google Tasks through per-user OAuth.
morning-digest
Assemble and deliver one short morning digest for the person who owns the cron — what changed overnight in the sources they already connected, and what it means for their day.
cron
Schedule reminders and recurring tasks.