babyworm/rtl-agent-team/rtl-p4s-bugfix-policy

Claude Code and agent skill by babyworm/rtl-agent-team

Trust tier 3
  • community
Source check pending

Checking that the SKILL.md file is still in the repository.

Internal reference: rtl p4s bugfix policy (agent-loaded; do not invoke).

Install

npx vskill@latest install babyworm/rtl-agent-team/rtl-p4s-bugfix-policy

SKILL.md

RTL Bug Fix Policy

Mandatory 4-Step Sequence (per module)

  1. Analyze → identify root cause and affected modules
  2. Fix + Lint → modify RTL, run verilator --lint-only -Wall (max 3 lint rounds)
  3. TB → create or update testbench (at least 1 test per modified module)
  4. Functional Verification → run simulation, ALL tests must PASS

Each step can only proceed after the previous step is complete. Lint is a necessary condition, NOT sufficient — simulation is required.

Parallel UNIT_FIX Decision Tree

Failure PatternHandlingParallelism
Different modules, UNIT_FIXParallel rtl-p4s-bugfix, one per modulerun_in_background: true
Same module, multiple failuresSequential within single taskNo
INTEGRATION_FIX (cross-module)Always sequentialNo
Mixed UNIT_FIX + INTEGRATION_FIXINTEGRATION_FIX first, then UNIT_FIX in parallelPartial

Verification-done marker created ONLY after ALL parallel fixes pass functional verification.

Phase 5→4 Feedback Mode

Fix TypeAfter FixRe-verify
UNIT_FIX (SVA fail)lint → unit TB → unit simOnly affected Phase 5 sub-phase
UNIT_FIX (sim fail)lint → unit TB → unit simOnly affected Phase 5 sub-phase
INTEGRATION_FIXlint → unit TB + integration TB → sim5b + 5c

In feedback mode: lesson-learned recording is mandatory (not just recommended).

Escalation & Stop Conditions

  • Lint fails 3 times → escalate to rtl-architect for design review
  • Cannot write TB (no ref model) → write minimal self-checking TB, report ref model requirement to user
  • Simulation fails after 2 fix iterations → escalate to waveform-analyzer + rtl-bug-repro skill
  • Simulator not installed → eda-runner provides installation instructions

TB Signal Naming Convention

  • dut.sys_clk, dut.sys_rst_n (clock/reset)
  • dut.i_*, dut.o_* (input/output ports)
  • TB check: ls sim/*/test_*.py sim/*/tb_*.sv 2>/dev/null

Advanced Rules

Integration bugs (multi-module):

  • Both per-module unit TBs AND top-level integration TB needed
  • Unit TBs: verify individual module I/O
  • Integration TB: verify data flow between modules

When regression suite available:

  • Recommended to re-run full regression after fix
  • Use rtl-p5s-func-verify skill for multi-seed verification

Compound mode bugs (e.g., MODE_RECON):

  • Test individual modes first, then compound mode tests, then mode transition scenarios

Final Checklist

  • [ ] Bug root cause identified
  • [ ] RTL fix complete
  • [ ] Lint passed (verilator --lint-only -Wall, 0 errors)
  • [ ] TB created or updated (at least 1 test per modified module)
  • [ ] Bug reproduction scenario included as a test case
  • [ ] Functional simulation executed (cocotb or verilator sim)
  • [ ] All tests PASS
  • [ ] Verification-done marker created (.rat/state/rtl-verify-done)
  • [ ] TB signal naming convention followed
  • [ ] Multi-module failures classified (UNIT_FIX vs INTEGRATION_FIX)
  • [ ] Independent UNIT_FIX modules fixed in parallel (not sequentially)
  • [ ] All parallel fixes passed functional verification before verify-done marker

Security

  • Security scanPassed, 90 of 100Pattern checks on the skill's files.
  • Model reviewPassed

Quality evaluation

  • Task evaluationNot runNo evaluation yet. Run one in Skill Studio.

Source

Popularity

Repository stars
61
Repository forks
16
7-day trend score
0

Stars and forks are for the whole GitHub repository, not this skill alone. Not refreshed yet.

Works with

Any agent vskill installs to: 56 targets in all. These 39 have a page on this site.

  • Claude Code
  • Cursor
  • GitHub Copilot
  • Windsurf
  • Codex
  • Gemini CLI
  • Cline
  • Amp
  • Roo Code
  • Goose
  • Continue
  • Kilo Code
  • Junie
  • Trae
  • OpenHands
  • Qwen Code
  • AI Agent
  • Tabnine Chat
  • Sourcegraph Cody
  • Amazon Q
  • Firebase Genkit
  • Sweep
  • Devin
  • Phind
  • Mentat
  • Claude Desktop
  • Aider
  • GPT Pilot
  • Plandex
  • bolt.new
  • v0
  • Replit Agent
  • Anse
  • TypingMind
  • BetterChatGPT
  • LibreChat
  • Warp AI
  • Pieces
  • Fig AI
Embed badge
vskill VERIFIED badge
[![vskill VERIFIED](https://verified-skill.com/api/v1/skills/babyworm/rtl-agent-team/rtl-p4s-bugfix-policy/badge)](https://verified-skill.com/skills/babyworm/rtl-agent-team/rtl-p4s-bugfix-policy)

Something wrong with this skill or its page? Tell us.