Help us improve
Share bugs, ideas, or general feedback.
From rtl-agent-team
Defines RTL performance verification policy: monitor naming conventions, BFM comparison thresholds (10% deviation), escalation rules, and stress vector patterns for cycle-accurate measurement.
npx claudepluginhub babyworm/rtl-agent-team --plugin rtl-agent-teamHow this skill is triggered — by the user, by Claude, or both
Slash command
/rtl-agent-team:rtl-p5s-perf-policyThe summary Claude sees in its skill listing — used to decide when to auto-load this skill
Performance monitor instrumentation and testbenches MUST follow project conventions (CLAUDE.md):
Measures RTL throughput, latency, and stall cycles against BFM-predicted baselines, flagging deviations over 10% and generating a PASS/FAIL performance report.
Guides technical evaluation of code review feedback: read fully, restate for understanding, verify against codebase, respond with reasoning or pushback before implementing.
Share bugs, ideas, or general feedback.
Performance monitor instrumentation and testbenches MUST follow project conventions (CLAUDE.md):
i_ prefix for inputs, o_ prefix for outputs (e.g., i_valid, o_stall)clk (single domain) or {domain}_clk (multiple domains, e.g., sys_clk) — NOT clk_irst_n (single domain) or {domain}_rst_n (multiple domains, e.g., sys_rst_n) — NOT rst_niu_ prefix (e.g., u_perf_counter)logic for all signal declarations (NOT reg/wire)i_/o_ prefix, sys_clk)Performance vectors should stress maximum throughput (back-to-back i_valid high, no gaps).
Also run a stall-stress vector (frequent o_ready deassertion) to expose stall handling bugs.
Use $time with sys_clk edges for cycle-accurate measurement in SV testbenches.