agentmods.dev

code-by-wire CLAUDE.md

Project instructions for code-by-wire, an application built with Electron, React, and TypeScript. They describe development commands, testing, formatting, native-module rebuilding, cross-platform requirements, and code organization.

In plain words
What is it for?
Use them when developing, testing, type-checking, formatting, or preparing changes for the code-by-wire application.
Why use it?
They give a coding agent the repository’s expected tools and checks, reducing mistakes such as using the wrong package manager or skipping required builds.

Instructions file

74 6d ago A scan 1,073 tokens original MIT
View the source file luojiahai/code-by-wire
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/luojiahai/code-by-wire/claude-md
Clone the repo
git clone --depth 1 https://github.com/luojiahai/code-by-wire
Per session 1,073 This file is loaded in full into every session.
When invoked 1,073 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 $0.01073 $0.01073
Opus 5 $0.00536 $0.00536
Sonnet 5 $0.00215 $0.00215
Haiku 4.5 $0.00107 $0.00107

Measured 2d ago against content hash 8188ca71f990, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

code-by-wire CLAUDE.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 2d 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.md · 51 lines

How it starts

The opening of the file, as written. The whole thing — 51 lines — stays where its author put it; the contents beside it link to each section on GitHub.

CLAUDE.md

This file provides guidance to Claude Code (claude.ai/code) when working with code in this repository.

Development

  • Package manager is pnpm (pinned), Node 24. Use pnpm, never npm.
  • pnpm dev runs the app (electron-vite). pnpm test runs the suite (vitest); tests live in tests/ mirroring the source tree. tests/renderer/** and tests/xterm/** run under jsdom (renderer logic only); there's no component-render harness, so verify actual UI by hand. Vitest only picks up tests/**/*.test.ts — no .tsx tests.
  • CI runs pnpm test on both ubuntu-latest and windows-latest (.github/workflows/ci.yml) — terminal/pty and path-handling code must work on both, not just macOS where local dev happens.
  • After pnpm install or any Electron upgrade, run pnpm rebuild:native — better-sqlite3 and node-pty are native modules and must be rebuilt against Electron's ABI, or the app crashes on launch.
  • pnpm typecheck runs two passes: tsconfig.node.json (main/preload/shared/tests) and tsconfig.web.json (React renderer, JSX). Test-reachable types must live in JSX-free .ts so they pass under the node config.
  • Before pushing, run pnpm format and pnpm lint — CI's lint job runs format:check then lint and fails on either.
  • scripts/make-icon.mjs (pnpm icon) string-replaces exact lines from build/icon.svg — editing the SVG can silently break macOS icon generation; check the script's .replace() targets after any SVG change.

Website

website/ (marketing site, Astro + Vercel) is a separate project sharing this repo — it has its own pnpm-workspace.yaml, isolating its install and lockfile from the root workspace. Run its commands from inside website/: pnpm install, pnpm dev, pnpm run check (typecheck), pnpm test.

Architecture

Electron app, three processes plus a utility process:

  • main (src/main/) — Node. Reads Claude Code and Codex transcripts (provider/claude/, provider/codex/), analytics in better-sqlite3 (db/), pty terminals (terminal/), git, settings. Request/response only — no background timers or fs.watch; the renderer polls.
  • parse worker (src/main/parse-worker/, its own bundle entry → out/main/parse-worker.js) — a utilityProcess main forks lazily so every poll-driven fs read (summarize, discovery, transcript/metrics/shells/monitors/tasks reads, the reconcile sweeps) runs off the main thread. It hosts the same ClaudeReader main falls back to; transcript docs return as one JSON string (TranscriptReadWire) that main relays and the preload parses. Still request/response (rides renderer polls); on worker fault each read degrades to the same code in-process. Main-process state (managed registry, model picks) never moves in — it rides the request or adorns the response.
  • preload (src/preload/) — contextBridge exposing window.api to the renderer.
  • renderer (src/renderer/src/) — React 19 + Tailwind 4 + xterm.

Read the full file on GitHub · 51 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. 2d ago First seen · 51 lines · 1,073 tokens per session scan A 8188ca71f990

Subscribe to this mod's changes

code-by-wire CLAUDE.md is an instructions file published in the GitHub repository luojiahai/code-by-wire (74 stars, last pushed 6d ago), licensed MIT. It adds 1,073 tokens to every session, about $0.0054 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 2026-08-30.

FrontendPlanningReActTestingclaude-codedesktop-appelectronterminaltypescript
Related

Other instructions, from other repositories

ClawUI AGENTS.md

Instructions for dawangcoding/ClawUI, covering clawui 开发基本指南, 项目概述, 技术栈, 编码规范 and 代码格式化 (prettier).

dawangcoding/ClawUI · 6,103 tokens

swob AGENTS.md

Instructions for IvyYang1999/swob, covering swob 研发协作唯一规范, 不可越过的边界, 角色与职责, 生命周期 and 分支、worktree 与提交.

IvyYang1999/swob · 1,764 tokens

clawsuite AGENTS.md

Instructions for outsourc-e/clawsuite, covering clawsuite agent roster, how aurora uses this file, routing table, aurora-coder — frontend specialist and aurora-daemon — backend specialist.

outsourc-e/clawsuite · 2,525 tokens

cc-mascot CLAUDE.md

Instructions for kazakago/cc-mascot, covering cc mascot - 技術ドキュメント, プロジェクト概要, コンセプト, 技術スタック and アーキテクチャ概要.

kazakago/cc-mascot · 1,347 tokens

plano CLAUDE.md

Instructions for zqkra/plano, covering claude.md, what plano is, commands, environment gotchas (windows — hit these during setup) and environment gotchas (linux / fedora kde — port-specific).

zqkra/plano · 4,552 tokens

eIsland AGENTS.md

AGENTS.md instructions for JNTMTMTM/eIsland, covering codegraph, 1. think before coding, 2. simplicity first, 3. surgical changes and 4. goal-driven execution.

JNTMTMTM/eIsland · 1,237 tokens