chromeclaw CLAUDE.md

Project instructions for ChromeClaw, a Chrome extension that adds an AI chat panel to the browser and supports multiple AI providers.

In plain words
What is it for?
Use them when an AI coding agent needs to understand, change, test, or navigate the ChromeClaw extension.
Why use it?
They explain the project's monorepo, which is one repository containing several related packages, along with its commands and architecture.

Instructions file

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/algopian/chromeclaw/claude-md
Clone the repo
git clone --depth 1 https://github.com/algopian/chromeclaw
Per session 2,901 This file is loaded in full into every session.
When invoked 2,901 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 $0.02901 $0.02901
Opus 5 $0.01451 $0.01451
Sonnet 5 $0.00580 $0.00580
Haiku 4.5 $0.00290 $0.00290

Measured 2d ago against content hash f9c15c655d4c, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

chromeclaw 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 · 199 lines

How it starts

The opening of the file, as written. The whole thing — 199 lines — stays where its author put it; the contents beside it link to each section on GitHub.

CLAUDE.md — ChromeClaw Extension

What is this project?

ChromeClaw is a Chrome extension that provides AI chat in the browser's side panel with multi-provider LLM support. Built with React 19, TypeScript, and pi-mono (@mariozechner/pi-ai + @mariozechner/pi-agent-core). Users add their own API keys — no login or proxy required.

Monorepo layout

Flat monorepo orchestrated with Turborepo (turbo.json).

chromeclaw/
├── chrome-extension/              # Background service worker, manifest
│   └── src/background/
│       ├── agents/                # Agent personas, model adapter, stream handler
│       ├── channels/              # Telegram + WhatsApp messaging bridges
│       ├── context/               # System prompt assembly, context compaction
│       ├── cron/                  # Scheduled task runner
│       ├── errors/                # Error handling
│       ├── logging/               # Logging utilities
│       ├── media-understanding/   # Speech-to-text, media transcription
│       ├── memory/                # BM25 + embedding hybrid search, memory journal
│       ├── tools/                 # All tool implementations
│       └── tts/                   # Text-to-speech (OpenAI, Kokoro)
├── pages/
│   ├── side-panel/                # Primary chat UI (overlay sidebar mode)
│   ├── full-page-chat/            # Full-page chat (push sidebar mode)
│   ├── offscreen-channels/        # Offscreen page for channel message handling
│   └── options/                   # Settings page (tabbed, see below)
├── packages/
│   ├── baileys/                   # WhatsApp (Baileys) integration
│   ├── config-panels/             # Options page tab panels and tab group definitions
│   ├── dev-utils/                 # Dev utilities
│   ├── env/                       # Build-time CEB_* environment variables
│   ├── hmr/                       # Hot module reload for extension dev
│   ├── i18n/                      # Internationalization
│   ├── module-manager/            # Module dependency management CLI
│   ├── shared/                    # Types, hooks (useLLMStream), prompts, env config
│   ├── skills/                    # Skill template loading and parsing
│   ├── storage/                   # Chrome storage + IndexedDB (Dexie.js) — all persistence
│   ├── tailwindcss-config/        # Tailwind configuration
│   ├── tsconfig/                  # Base TypeScript configs
│   ├── ui/                        # React components (shadcn/ui + custom chat components)
│   ├── vite-config/               # Shared Vite configuration
│   └── zipper/                    # Extension ZIP packaging
├── tests/playwright/              # E2E tests
├── bash-scripts/                  # Shell scripts (copy-env, set-global-env, update-version)
├── docs/                          # Documentation
├── turbo.json                     # Turborepo config
├── vitest.config.ts               # Shared Vitest config
└── playwright.config.ts           # Playwright config

Read the full file on GitHub · 199 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 · 199 lines · 2,901 tokens per session scan A f9c15c655d4c

Subscribe to this mod's changes

chromeclaw CLAUDE.md is an instructions file published in the GitHub repository algopian/chromeclaw (196 stars, last pushed 1mo ago), licensed MIT. It adds 2,901 tokens to every session, about $0.0145 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.