hernanjiang/craftstation/chrome-control

Chrome Control

Use CraftStation's chrome MCP when a task depends on the user's current Chrome tabs, authenticated sessions, or installed extensions. If a purpose-built connector can complete a semantic service operation, prefer it unless the user explicitly requested Chrome or visual interaction is part of the task.

Workflow

1. Call chrome.chrome_status first. If the extension is disconnected, ask the user to connect it rather than switching surfaces silently.
2. Call chrome.enable once before browser actions. Use the background CraftStation workspace by default; call chrome.chrome_attach only when the user asked to operate an existing tab.
3. Inspect with chrome.chrome_snapshot or chrome.chrome_find before clicking or typing. Prefer returned element refs and use chrome_fill for replacement versus chrome_type for appending.
4. Keep every action scoped to the requested site and task. Do not explore other tabs or signed-in content for extra context.
5. After every meaningful action, wait for and verify the resulting URL, text, control state, or screenshot.
6. Call chrome.disable before asking the user for input, waiting on an external event, or finishing.

Boundaries

• Treat tabs, cookies, storage, and signed-in content as sensitive user data. Do not read cookies unless the task requires it and Chrome data access is enabled.
• Never attach to an unrelated existing tab merely because it is already authenticated.
• A successful tool call is not proof that the website accepted the action; verify the visible result.
• Confirm the exact target and payload before purchases, submissions, messages, deletions, account changes, or other irreversible actions unless already authorized.

Output

Report the target site or tab, what changed, and the final visible evidence. State whether the background workspace or an existing user tab was used, and identify any action left pending for confirmation.