What is opencode-webmcp AGENTS.md?

A set of project instructions for a Chromium WebMCP bridge. WebMCP is a Chromium feature that lets webpages register tools for software agents to use.

Use it as a reference when starting the server, selecting Chrome or Chromium, setting the initial webpage, and understanding the bridge's limited runtime verification.

It gives an agent the project facts needed to run and modify the bridge correctly, including its Chrome version requirement and environment settings.

opencode-webmcp AGENTS.md — Instructions file by enihcam

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/enihcam/opencode-webmcp/agents-md

Clone the repo

git clone --depth 1 https://github.com/enihcam/opencode-webmcp

Made for: Codex, OpenCode.

Per session 1,900 This file is loaded in full into every session.

When invoked 1,900 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.

Model	Per session	Once invoked
Fable 5	$0.01900	$0.01900
Opus 5	$0.00950	$0.00950
Sonnet 5	$0.00380	$0.00380
Haiku 4.5	$0.00190	$0.00190

Measured yesterday against content hash f3be1fa6b960, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

opencode-webmcp 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 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.

AGENTS.md · 136 lines

How it starts

The opening of the file, as written. The whole thing — 136 lines — stays where its author put it; the contents beside it link to each section on GitHub.

webmcp-bridge

Single-file MCP server that bridges MCP clients (OpenCode, Claude Code) to Chromium's WebMCP API. Connects via Puppeteer + CDP, discovers tools registered via document.modelContext on the page, and exposes them as MCP tools.

Key facts

Entry point: server.js (ESM, Node 18+, ~400 lines)
Deps: @modelcontextprotocol/sdk + puppeteer-core
No build step, no test suite, no linter, no typecheck, no npm scripts
Only runtime verification: node server.js (needs Chrome 150+)

Chrome requirement

puppeteer-core — no bundled browser. Must have system Chrome/Chromium 150+. Set CHROME_PATH (default /usr/bin/chromium). WebMCP gated behind --enable-features=WebMCPTesting,DevToolsWebMCPSupport.

Running

CHROME_PATH=/usr/bin/chromium WEBMCP_TARGET_URL="about:blank" WEBMCP_HEADLESS=false node server.js

Env var	Default	Notes
`CHROME_PATH`	`/usr/bin/chromium`	Path to Chrome/Chromium 150+
`WEBMCP_TARGET_URL`	`https://www.google.com`	Initial page to navigate to
`WEBMCP_HEADLESS`	`true`	`false` for visible window

--no-sandbox is always set; --ozone-platform=wayland added when not headless.

Architecture

Single file, no build — edit server.js, restart to apply changes.
5 bridge-native tools: webmcp_navigate, webmcp_status, webmcp_evaluate, webmcp_invoke_tool, webmcp_register_test_tools — always available.
Page-discovered WebMCP tools served alongside bridge tools.
Monkey-patch (evaluateOnNewDocument): Chrome 150+ has a bug where page.webmcp.invokeTool() returns {}. The bridge intercepts document.modelContext.registerTool to capture each tool's execute function, then replaces executeTool to call captured functions directly.
Auto-recovery: detached-frame errors re-create the page automatically. Bridge-native tools retry; page-discovered tool calls re-throw with a retry instruction.
notifications/tools/list_changed sent on navigation and tool registration changes.

Read the full file on GitHub · 136 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.

yesterday First seen · 136 lines · 1,900 tokens per session scan A f3be1fa6b960

Subscribe to this mod's changes

opencode-webmcp AGENTS.md is an instructions file published in the GitHub repository enihcam/opencode-webmcp (1 stars, last pushed 8d ago), licensed MIT. It adds 1,900 tokens to every session, about $0.0095 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.

Browser automation Planning Productivity ai-tools browser-automation chrome-devtools chrome-extension claude-code mcp

opencode-webmcp AGENTS.md

Getting it into your agent

What it costs to keep this loaded

Grade A, and why

Nothing flagged

How it starts

webmcp-bridge

Key facts

Chrome requirement

Running

Architecture

What this file has done since we first saw it

Other instructions, from other repositories

codex-browser-bridge AGENTS.md

mcp-browser-bridge CLAUDE.md

grok-chrome-bridge AGENTS.md

AgentLimb AGENTS.md

AgentLimb CLAUDE.md

browser-use CLAUDE.md