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.
npx agentmods add instructions/saladtechnologies/mcproxy/claude-mdgit clone --depth 1 https://github.com/SaladTechnologies/mcproxyWhat 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.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5 | $0.02321 | $0.02321 |
| Opus 5 | $0.01161 | $0.01161 |
| Sonnet 5 | $0.00464 | $0.00464 |
| Haiku 4.5 | $0.00232 | $0.00232 |
Grade A, and why
mcproxy 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 yesterday.
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.
How it starts
The opening of the file, as written. The whole thing — 257 lines — stays where its author put it; the contents beside it link to each section on GitHub.
MCProxy - Claude Code Context
Project Overview
MCProxy is a remote browser automation system via MCP (Model Context Protocol). It consists of two main components:
- Browser Server (
browser-server/): Runs on SaladCloud containers, provides headless browser automation via WebSocket - MCP Server (
mcp-server/): Runs locally, exposes browser tools to AI agents via MCP stdio protocol
Architecture
Agent (Claude) → MCP Server (local, stdio) → WebSocket → Browser Server (SaladCloud) → Playwright
- MCP Server manages multiple WebSocket connections to different browser server instances
- Each browser server can host multiple browser contexts (sessions)
- Sessions are location-aware and report geographic info (IP, city, region, country)
Monorepo Structure
mcproxy/
├── shared/ # Shared TypeScript types (@mcproxy/shared)
│ └── src/protocol.ts # WebSocket message types, command params, capabilities
├── browser-server/ # Remote browser server (@mcproxy/browser-server)
│ ├── src/
│ │ ├── browser-manager.ts # Playwright browser/context management, stealth
│ │ ├── capabilities.ts # Server capability reporting
│ │ ├── captcha-detector.ts # CAPTCHA detection logic
│ │ ├── command-handler.ts # Command execution (navigate, click, etc.)
│ │ ├── location-service.ts # IP geolocation detection
│ │ └── ws-server.ts # WebSocket server with auth
│ └── Dockerfile
├── mcp-server/ # Local MCP server (@mcproxy/mcp-server)
│ ├── src/
│ │ ├── browser-client.ts # WebSocket client with heartbeat
│ │ ├── credential-store.ts # Secure credential storage and scrubbing
│ │ ├── session-manager.ts # Session lifecycle management
│ │ └── tools/index.ts # MCP tool definitions and handlers
│ └── Dockerfile
└── .github/workflows/ # CI/CD for Docker images
Development Commands
# Install dependencies
npm install
# Build all packages (shared must build first)
npm run build
# Build specific package
npm run build:shared
npm run build:browser-server
npm run build:mcp-server
# Development with hot reload
npm run dev:browser-server
npm run dev:mcp-server
# Local testing with Docker
docker compose up --build
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.
- yesterday First seen · 257 lines · 2,321 tokens per session scan A 4d3caa02292d
mcproxy CLAUDE.md is an instructions file published in the GitHub repository SaladTechnologies/mcproxy (2 stars, last pushed 7mo ago), licensed MIT. It adds 2,321 tokens to every session, about $0.0116 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.
Other instructions, from other repositories
Browser4 CLAUDE.md
Instructions for platonai/Browser4, covering browser4 — project context for claude, architecture, key dispatch chain (cli → browser), batch commands and e2e test structure.
spring-ai-agentcore AGENTS.md
Instructions for spring-ai-community/spring-ai-agentcore, covering agents.md, project overview, architecture, key components and artifact store classes.
flyto-core CLAUDE.md
Instructions for flytohub/flyto-core, covering claude notes, cross-agent handoff and shared code intelligence.
deckforge AGENTS.md
Instructions for tph-kds/deckforge, covering deckforge agent entry point, code intelligence, read order, default routing and non-negotiable implementation rules.
fast-browser CLAUDE.md
Claude Code instructions for m4ttstack/fast-browser, covering fast browser plugin, where a change belongs, fork branch: use fast-browser-runtime, releasing a new runtime and re-pinning this repo: use the script.
agentoscope AGENTS.md
Instructions for rafaelcg/agentoscope, covering agentoscope — agent instructions, what this is, layout, commands and conventions.