KiCadRoutingTools: Skill for Claude Code

.claude/skills/review-routed-board/SKILL.md

review-routed-board is a skill for Claude Code from drandyhaas/KiCadRoutingTools. It costs 83 tokens per session (5,630 once invoked), scanned A, original, MIT.

A post-routing quality check for a KiCad circuit-board design. It checks electrical rules, connectivity, leftover route stubs, unusual copper or vias, timing and length groups, ground return coverage, and differential pairs.

In plain words
What is it for?
Use it to review a routed board, check high-speed signal constraints, verify differential pairs and ground paths, and prepare a final sign-off report.
Why use it?
It finds routing problems that may remain after automatic or manual routing and produces a pass/fail report with next steps.

Skill for Claude Code

Written for Claude Code: installed under .claude/.

This is drandyhaas/KiCadRoutingTools's own configuration. It tells Claude Code how to work on KiCadRoutingTools itself, so it is not a mod to install elsewhere. Copy it as a starting point and replace the rules that are about this project. Everything KiCadRoutingTools configures →

Needs its repository: it runs a file that does not travel with it, so clone the repository first. The line is KCLI=$(python3 py_tools/kicad_cli_path.py).

Reuse

Borrowing it

Nothing to install: this file belongs to drandyhaas/KiCadRoutingTools. Take a copy, put it at the same path in your own repository, and replace the rules that are about this project with yours.

Copy the file
curl -O https://raw.githubusercontent.com/drandyhaas/KiCadRoutingTools/main/.claude/skills/review-routed-board/SKILL.md
Clone the repo
git clone --depth 1 https://github.com/drandyhaas/KiCadRoutingTools

Made for: Claude Code.

Wrote 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.

agentmods badge for review-routed-board

README.md
[![agentmods](https://agentmods.dev/badge/skills/drandyhaas/kicadroutingtools/review-routed-board/github.svg)](https://agentmods.dev/skills/drandyhaas/kicadroutingtools/review-routed-board)
Your own site
<a href="https://agentmods.dev/skills/drandyhaas/kicadroutingtools/review-routed-board"><img src="https://agentmods.dev/badge/skills/drandyhaas/kicadroutingtools/review-routed-board/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.

agentmods 80×15 button for review-routed-board

Your own site · 80×15
<a href="https://agentmods.dev/skills/drandyhaas/kicadroutingtools/review-routed-board"><img src="https://agentmods.dev/badge/skills/drandyhaas/kicadroutingtools/review-routed-board.svg" alt="Reviewed on agentmods" width="80" height="20"></a>
Per session 83 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 5,630 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. A grade says what 26 rules found in the file — not that it is safe. Third-party audits
  • NVIDIA SkillSpector pass 7 Sept 2026
How audits are shown
Origin original No closer match found in the catalogue.
Token cost

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.

ModelPer sessionOnce invoked
Fable 5.1 $0.00083 $0.05630
Opus 5.5 $0.00033 $0.02252
Sonnet 5.5 $0.00017 $0.01126
Haiku 4.5 $0.00008 $0.00563

Measured 3d ago against content hash 748ba75dd748, method: parsed. Prices are Anthropic first-party input rates as of 2026-10-07, from the pricing page.

Security

Grade A, and why

review-routed-board scanned grade A with 0 findings 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 3d 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.

Nothing flagged

None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.

.claude/skills/review-routed-board/SKILL.md · 381 lines

How it starts

The opening of the file, as written. The whole thing — 381 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Review Routed Board

When this skill is invoked with a board file, run a full post-route review and present a sign-off report.

Step 1: Mechanical Checks

Run all four checkers, capturing output:

python3 -X utf8 py_router/check_drc.py board.kicad_pcb 2>&1 | tee /tmp/review_drc.txt
python3 -X utf8 py_router/check_connected.py board.kicad_pcb 2>&1 | tee /tmp/review_connectivity.txt
python3 -X utf8 py_tools/check_orphan_stubs.py board.kicad_pcb 2>&1 | tee /tmp/review_orphans.txt
python3 -X utf8 py_router/check_weird.py board.kicad_pcb 2>&1 | tee /tmp/review_weird.txt

check_weird.py is here because the other three cannot see its classes. check_orphan_stubs iterates SEGMENT endpoints and treats a via as an anchor, so it structurally cannot report a bad via; check_weird owns dangling-via (same-net copper on only one of the layers the barrel spans -- KiCad's via_dangling), unsupported-via, stacked-copper, orphan-island and kicad-dangling (a joint stub lying on one other track, which KiCad reports as track_dangling though both its ends touch copper). Measured on run 11's final board: check_orphan_stubs none, check_weird 3 dangling vias, each independently confirmed.

check_drc.py auto-grades at the clearance the routing steps wrote into the sibling .kicad_pro (the smallest clearance any step actually used, including auto-stepped fine-pitch taps), so the bare invocation above already grades at the true routed floor. Like KiCad, it floors that class at Board Setup min_clearance (#1210): an unrouted or hand-routed board whose project declares a minimum above its Default class grades at the minimum. Pass --clearance <value> only to override (e.g. to grade a hand-routed board with no routed-floor .kicad_pro).

When the board was routed from an input you have, add --baseline <the input board>. VIA-IN-PASTE rows are vias whose barrel sits in a solder-paste opening of their own net without IPC-4761 Type VII (filled AND capped); solder wicks into such a barrel. KiCad has no such check, so check_drc is the only instrument that reports them. On a KiCad 10-format board the router stamps Type VII onto every via it adds in a pad or paste opening, so these are almost always vias the input already had: --baseline accepts those (a via the input had under solder, unprotected) as inherited-via-in-paste and the console line counts them. A file older than KiCad 10 (version < 20250000) cannot carry the tokens at all, so there every such via --baseline does not inherit is accepted undeclarable-via-in-paste and counted on the same line: Type VII belongs on the fab drawing. Report whatever still fires as a fab defect: the via needs filled+capped on the fab drawing, or it must move out of the opening. Without --baseline, every pre-existing one reads as a violation (one corpus board carries 136). The same flag grades a graze of footprint graphic copper that a part MOVE created (graphic-board-edge); graphic-off-board (copper past the outline) is reported either way. KEEPOUT rows are tracks or vias inside a keep-out rule area that forbids them (an antenna clearance, a mounting hole's screw-head ring); kicad-cli 10.0.0 did not report these when measured, so check_drc is the only instrument. --baseline accepts copper the input already had there as inherited-keepout; report what still fires as a routing defect.

Read the full file on GitHub · 381 lines

Changes

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.

  1. 3d ago Changed · +6 lines 748ba75dd748
  2. 4d ago Changed · +1 lines 3fd28e59fbec
  3. 5d ago Changed · +18 lines b44a8970467d
  4. 6d ago Changed · +2 lines 174e46920b1d
  5. 7d ago Changed · +40 lines 8ac5d69d4fbe
  6. 13d ago Changed · +12 lines 29e7c0e3a11e
  7. 18d ago Changed · +4 lines 5a89d48722ba
  8. 20d ago Changed · +8 lines 63117d414c5a
  9. 26d ago Changed · +2 lines b6b8bdf9bcb9
  10. 1mo ago First seen · 288 lines · 83 tokens per session scan A b45c742a9f1d

Subscribe to this mod's changes

review-routed-board is a skill published in the GitHub repository drandyhaas/KiCadRoutingTools (706 stars, last pushed today), licensed MIT. It adds 83 tokens to every session and 5,630 once invoked, about $0.0003 per session on Opus 5.5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.

Related

Other skills, from other repositories