Use a headed browser to automatically complete DingTalk channel integration for QwenPaw. Applicable when the user mentions DingTalk, developer console, Client ID, Client Secret, bot, Stream mode, binding or configuring a channel. Supports pausing when a login page is detected and resuming after the user logs in.
Use a headed browser to automatically complete DingTalk channel integration for QwenPaw. Applicable when the user mentions DingTalk, developer console, Client ID, Client Secret, bot, Stream mode, binding or configuring a channel. Supports pausing when a login page is detected and resuming after the user logs in.
This skill automates the creation of a DingTalk application and the binding of a QwenPaw channel using a headed browser.
Mandatory Rules
Must launch in headed browser mode:
{"action":"start","headed":true}
Must pause when a login gate is encountered:
If the page displays a login screen (e.g., login prompt, QR code login, phone/password login), stop automated operations immediately.
Clearly prompt the user to log in manually first, then wait for the user to reply with "logged in / continue".
Do not proceed with subsequent steps until the user confirms.
Any application configuration change only takes effect after creating a new version and publishing:
After configuring bot-related information, .
you must publish the bot
Whether creating a new application or modifying application information (name, description, icon, bot configuration, etc.), you must perform "create new version + publish" at the end.
Do not claim the configuration is active if publishing has not been completed.
Pre-Execution Confirmation (Must Do First)
Before starting automated clicks, initiate a "configuration confirmation" with the user, clearly informing them of customizable fields, image specifications, and default values. Use the following structured confirmation:
Allow the user to customize the following fields:
Application name
Application description
Bot icon image URL or local path
Bot message preview image URL or local path
Clearly state the image specifications (prominently):
Bot icon: JPG/PNG only, 240*240px or larger, 1:1 ratio, under 2MB, no rounded corners.
Bot message preview image: format png/jpeg/jpg, no more than 2MB.
Clearly state the default values (used automatically if the user does not specify):
If the page is within an iframe, prefer including frame_selector; otherwise the upload control may not be found or the chooser may not trigger.
You must click the upload entry before calling file_upload; calling it directly will result in:
No chooser. Click upload then file_upload.
Common structural features of the bot icon area that can be used for locating elements (examples; these may appear as Chinese UI labels in the DingTalk console):
When the snapshot shows both "使用应用图标" ("Use App Icon") and "avatar", prefer clicking the avatar button to trigger the upload, then call file_upload.
Automation Flow
Step 1: Open the DingTalk Developer Console
Launch the browser in headed mode (headed: true)
Navigate to https://open-dev.dingtalk.com/
Call snapshot to check if login is required
If login is required, pause with the following message:
Login to the DingTalk Developer Console is required. I have paused automated operations. Please complete the login in the opened browser. Reply "continue" when done, and I will resume from the current page.
If it is a link, download to local first, then upload
If the image does not meet specifications, pause and ask the user to manually upload a compliant image before continuing
Confirm that the message receiving mode is set to Stream Mode (the Chinese UI may display Stream 模式)
Select Publish; a further confirmation dialog will appear -- select publish. Note: you must publish the bot before proceeding to the next step
Step 4: Create Version and Publish
Navigate to Application Release -> Version Management & Release
Create a new version (required after every configuration change)
Fill in the version description; set the application visibility scope to all employees
Follow the page prompts to complete publishing; a new dialog will appear -- select confirm publish
Only after seeing the successful publication status may you proceed with subsequent steps or tell the user "the configuration is now active"
Step 5: Obtain Credentials
Navigate to Basic Information -> Credentials & Basic Info
Inform the user that the Client ID (AppKey) and Client Secret (AppSecret) are on this page. Do not make changes proactively; guide the user to bind them on their own
QwenPaw Binding Methods
After obtaining the credentials, guide the user to choose one of the following methods:
Console frontend configuration:
In the QwenPaw console, go to Control -> Channels -> DingTalk
Path: ~/.qwenpaw/config.json, under channels.dingtalk.
Credential Delivery Requirements (Mandatory)
The agent is only responsible for guiding the user to the credentials page, obtaining and displaying the Client ID and the actual Client Secret.
The agent must not proactively modify the console configuration or ~/.qwenpaw/config.json.
You must instruct the user to fill in the credentials manually using one of the following two methods:
Console frontend: Control -> Channels -> DingTalk
Configuration file: edit the channels.dingtalk field in ~/.qwenpaw/config.json
Browser Tool Usage Pattern
Execute in the following order by default:
start with headed: true
open
snapshot
click / type / select_option / press_key as needed
frequent snapshot after page transitions
stop when done
Stability and Recovery Strategy
Prefer using the ref from the latest snapshot; only use selector when necessary.
After each critical click or navigation, use a short wait (wait_for) and immediately re-run snapshot.
If the session expires or re-login is required mid-flow, pause again and wait for the user to log in before continuing from the current step.
If automation is blocked by tenant permissions or admin approval, clearly describe the blocker and ask the user to manually complete that step before resuming.
