kicad-mcp AGENTS.md

kicad-mcp AGENTS.md is an instructions file for Codex, OpenCode from blwfish/kicad-mcp. It costs 1,386 tokens per session, scanned A, original, MIT.

Repository instructions for kicad-mcp, a tool that connects an AI agent with KiCad, software for designing circuit boards. They explain how to keep circuit-board decision logic testable when scripts run across the KiCad process boundary.

In plain words
What is it for?
Use them when adding or editing KiCad tools, extracting helper logic, writing tests that do not require KiCad, and following the project's contribution rules.
Why use it?
They prevent important geometry and classification logic from being buried inside hard-to-test KiCad scripts.

Instructions file for CodexOpenCode

Install

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.

agentmods
npx agentmods add instructions/blwfish/kicad-mcp/agents-md
Clone the repo
git clone --depth 1 https://github.com/blwfish/kicad-mcp

Made for: Codex, OpenCode.

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 kicad-mcp AGENTS.md

README.md
[![agentmods](https://agentmods.dev/badge/instructions/blwfish/kicad-mcp/agents-md.svg)](https://agentmods.dev/instructions/blwfish/kicad-mcp/agents-md)
Your own site
<a href="https://agentmods.dev/instructions/blwfish/kicad-mcp/agents-md"><img src="https://agentmods.dev/badge/instructions/blwfish/kicad-mcp/agents-md.svg" alt="Measured on agentmods" height="20"></a>
Per session 1,386 This file is loaded in full into every session.
When invoked 1,386 The same file — it is already loaded in full.
Security scan A 0 findings. Scan, not verified.
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 $0.01386 $0.01386
Opus 5 $0.00693 $0.00693
Sonnet 5 $0.00277 $0.00277
Haiku 4.5 $0.00139 $0.00139

Measured 5d ago against content hash f9dbedc6f49f, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-05, from the pricing page.

Security

Grade A, and why

kicad-mcp AGENTS.md 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 5d 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.

AGENTS.md · 104 lines

How it starts

The opening of the file, as written. The whole thing — 104 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Agent working notes — kicad-mcp

Conventions for any AI agent (Claude Code, Cursor, etc.) working in this repo. General contributor guidance — bug reports, PR style, threshold-test coverage — is in CONTRIBUTING.md. This file holds the few rules that are easy to violate while generating code and expensive to retrofit.

Boundary ops: extract decision logic before welding it to pcbnew

When you add or edit a tool that runs a pcbnew or kicad-cli script — anything that builds a script string for run_pcbnew_script — do not put decision logic (geometry, thresholds, classification) inside that string. Across the process boundary it becomes unreachable to any in-process test, and the only test left is a tautological boundary-mock that asserts its own configured return and cannot fail.

Instead:

  1. Put the pure logic in a *_HELPER string in utils/ (exemplar: utils/geometry.pyGEOMETRY_HELPER).
  2. Keep the embedded script a thin shell: load the board, marshal pcbnew objects into plain dicts, splice the helper in by concatenation (""" + _MY_HELPER + """), and call the helper for every decision.
  3. Add a no-KiCad test that execs the helper and asserts at the boundaries (value / just-below / just-above).

Full pattern and a copyable skeleton: docs/BOUNDARY_OPS.md. Straight-line marshalling (load → mutate → save, no branching) needs no helper — don't invent one.

Before finishing such work, run the report-only ratchet:

python scripts/audit_testability.py        # report-only; --check fails on NEW violations

It flags new helperless boundary logic and new tautological boundary tests. Pre-existing violations are grandfathered in scripts/testability_baseline.json; extracting one ratchets the count down. Refresh after an intentional change with --update-baseline.

Report which guarantee you hold — don't let "verified E2E" stand in for "tested"

When you report a change as done, name which guarantee you actually have. They are different, and a green golden-harness run is only the first:

Read the full file on GitHub · 104 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. 5d ago First seen · 104 lines · 1,386 tokens per session scan A f9dbedc6f49f

Subscribe to this mod's changes

kicad-mcp AGENTS.md is an instructions file published in the GitHub repository blwfish/kicad-mcp (12 stars, last pushed 5d ago), licensed MIT. It adds 1,386 tokens to every session, about $0.0069 per session on Opus 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 2026-08-31.