Chrome Web Automation

Use when the user wants browser automation in an existing Chrome session: click, fill, inspect tabs, capture screenshots, or debug web flows.

Security Scan

The skill claims to drive a user's live Chrome session (inspect tabs, click, type, screenshot) but provides no install steps, connector, or declared permissions for how it would access the browser — the claimed capability doesn't match the declared footprint.

Purpose & Capability

The name/description ask to control an existing Chrome session (inspect tabs, click, fill, screenshot). However the skill declares no binaries, no required env vars, no config paths, and has no code files. A skill performing live browser automation normally needs a connector (Chrome DevTools endpoint, extension, remote-debugging port, webdriver, or a platform-provided bridge) and explicit configuration. That capability is not declared here, making the required access unclear and disproportionate to what's stated.

Instruction Scope

SKILL.md tells the agent to inspect tabs, snapshot pages, click/type/submit and re-snapshot — actions that can access DOM, cookies, session state, and other sensitive data. The instructions are high-level and do not specify safe limits, exact APIs, or where data may be sent. Because the runtime mechanism is unspecified, the instructions grant broad discretion to interact with a user's active browser session, which is a privacy and scope-risk.

Install Mechanism

There is no install spec and no code files — this is instruction-only. That minimizes on-disk installation risk. However, the lack of an install or connector also contributes to the incoherence (no explanation of how browser access will be achieved).

Credentials

The skill declares no environment variables or credentials, yet the functionality would normally require explicit access tokens or endpoints (e.g., remote debugging port, extension token, socket path). The absence of declared credentials/configuration means it's unclear what privileges or secrets the skill will need or use, which is disproportionate and suspicious.

Persistence & Privilege

always is false and there's no install spec that writes persistent files or modifies other skills. Autonomous invocation is enabled by default (normal). While autonomous invocation combined with broad browser access would increase risk, there's no direct evidence here that the skill requests persistent elevated privileges.

What to consider before installing

This skill claims to control your live Chrome session but gives no details about how it will connect or what it will access. Before installing: ask the publisher how the agent connects to Chrome (DevTools websocket, extension, local socket, or platform bridge), request an explicit install/permission guide and a list of required environment variables or ports, and insist on minimal-scoped access. Do not grant autonomous invocation or broad browser access until you can review the connector implementation or source code. Treat this as potentially privacy-sensitive — browser automation can read cookies, sessions, and page contents; only proceed if you trust the skill's provenance and can audit the connection method.

Use this skill before any browser interaction that depends on the user's live Chrome session.

Workflow

1. Start by inspecting current tabs and browser context.
2. Navigate or select the right tab.
3. Snapshot the page before interacting.
4. Click, type, submit, and screenshot using stable references.
5. Re-snapshot after navigation or large DOM updates.

Best Fits

• Web app debugging
• Form filling
• Screenshot capture
• Console or tab inspection
• Reproducing UI issues in a user browser session

Guardrails

• Re-snapshot when element references go stale.
• Do not assume current tab state without checking.
• Prefer explicit browser actions over ambiguous natural-language leaps.

Source Provenance

Derived from src/skills/bundled/claudeInChrome.ts.