Cuchd Login

intent

automate full browser login to the openclaw platform and execute post-login tasks without confirmation at each step. use this skill when the user wants to log in, navigate the portal, submit forms, retrieve data, or interact with openclaw via browser. the skill grants autonomous control over clicking, typing, waiting, reading page content, and navigation. pause only on blockers like captchas or unexpected errors.

inputs

• openclaw username: employee id, user id, username, or email registered on the platform. passed as a string parameter.
• openclaw password: account password. never logged, stored, or echoed after use. passed as a secure parameter.
• browser session: agent has full browser access with cookies and session storage enabled.
• internet connectivity: reliable connection to https://openclaw.in required. network timeouts will fail the login.
• target url: https://openclaw.in (hardcoded). no environment variable setup required.

procedure

1. navigate to login page

   • input: target url https://openclaw.in
   • action: browser navigates to the url
   • output: openclaw login page html loaded in browser
   • validation: wait up to 15 seconds for page title and username field to appear. if blank page or network error after 15s, report timeout and stop.

2. locate and focus username field

   • input: rendered page dom
   • action: identify the username input field. may be labelled "username", "user id", "employee id", or "email". if field is already focused (cursor visible), skip to step 3. if not, click the field once.
   • output: username field is active and ready for text input
   • validation: confirm field has focus (border highlight or cursor visible).

3. enter username

   • input: username string from user
   • action: type the username into the focused field
   • output: username text appears in the field
   • validation: read field value back to confirm entry. if value is empty or truncated, report and stop.

4. submit username

   • input: current page state with username entered
   • action: locate and click the "next", "continue", or "proceed" button. if no button exists after scrolling, press the enter key instead.
   • output: page transitions or reloads
   • validation: wait up to 10 seconds for navigation to complete. if page still shows username field and no new elements appear, report "no next button found and enter key had no effect" and stop.

5. wait for password field

   • input: page state after username submission
   • action: wait up to 15 seconds for the password input field to become visible and interactable. the page may reload, dynamically update, or transition to a new view. do not assume the field is present until confirmed.
   • output: password field is visible in the dom and interactable
   • validation: scroll the page if the field is below the fold. if password field does not appear after 15s, report "password field did not appear" and describe visible ui elements.

6. enter password

   • input: password string from user, password field is active
   • action: click the password field if not already focused, then type the password. do not echo or log the password text.
   • output: password characters appear in the field (displayed as dots or asterisks for security)
   • validation: confirm the field is not empty by checking its character count. do not read the actual password value.

7. submit login

   • input: current page state with password entered
   • action: locate and click the "login", "sign in", or "submit" button. if no button exists after scrolling, press the enter key instead.
   • output: form is submitted, page navigates
   • validation: wait up to 15 seconds for navigation to complete and for the dashboard, home page, or next authenticated view to load.

8. confirm login success

   • input: page state after login form submission
   • action: read the page title, url, and any visible welcome or success message. check for error banners, alerts, or captcha elements.
   • output: login status report with page title, url, and any error or success message
   • validation: if error message is visible, extract and report the exact text. if captcha is present, stop and notify user. if dashboard or authenticated page title is visible, report success.

decision points

• if username field is auto-focused on page load: type directly without clicking. else click the field first to focus it.
• if a next/continue button exists: click it to proceed. else press enter after typing the username.
• if password field appears after username submission: proceed to step 6. else if page shows an error message, report the error and stop. else if page is blank or loading beyond 15s, report timeout.
• if login form is submitted and dashboard/home page loads: report success and username. else if an error message is visible on the page, report the exact error text and stop.
• if a captcha appears at any point: stop immediately and notify the user. do not attempt to solve it.
• if session expires during a post-login task: re-run the full login flow from step 1 automatically.
• if an unexpected popup or modal appears (e.g., cookies consent, update notification): dismiss if safe (e.g., close button visible), else report and pause.
• if a redirect occurs to an unknown or suspicious url: pause and report the new url to the user instead of following automatically.

output contract

on successful login, the skill outputs:

{
  "status": "login_successful",
  "username": "<entered username>",
  "page_title": "<title of authenticated page>",
  "current_url": "<url after login>",
  "timestamp": "<iso 8601 timestamp>"
}

on login failure, the skill outputs:

{
  "status": "login_failed",
  "error_message": "<exact text from error banner or alert>",
  "page_title": "<title of page showing error>",
  "current_url": "<url at time of error>",
  "timestamp": "<iso 8601 timestamp>"
}

on blocker (captcha, timeout, etc.), the skill outputs:

{
  "status": "blocker_encountered",
  "blocker_type": "captcha" | "timeout" | "element_not_found" | "network_error",
  "description": "<human readable description>",
  "current_url": "<url when blocker occurred>",
  "visible_ui": "<description of visible page content>",
  "timestamp": "<iso 8601 timestamp>"
}

all outputs are json and printed to stdout. no credentials are included in output.

outcome signal

the user knows the skill worked when:

• the browser displays the openclaw dashboard, home page, or authenticated portal (title and url indicate successful login).
• the user sees a welcome message, user profile, or navigation menu specific to authenticated sessions.
• if a post-login task was requested (e.g., "go to reports"), the agent has navigated to that section and reports the page title and url.
• the skill outputs a json object with "status": "login_successful" and includes the current page title and url.

the skill failed if:

• the login page is still visible after step 8, with no error message or captcha detected (ambiguous failure).
• a json object with "status": "login_failed" is output, including the exact error message from the page.
• a json object with "status": "blocker_encountered" is output (user intervention required).
• no output is produced after 60 seconds (network or browser failure).

original source: clawhub. original author not declared. enriched to implexa quality standards with explicit decision points, edge cases, error handling, and structured output contract.