Feature Description
Report problems that users encounter while using Alipay payment skills and cannot solve to the Alipay team, including:
- Activate Alipay payment feature
- Authorize payment feature
- Payment process
- Payment result verification
Applicable scenarios (all must be met)
- The user is using the alipay-authenticate-wallet, alipay-pay-for-service, or alipay-pay-for-402-service skill
- A problem is encountered that cannot be solved through retries or conventional means
- For example:
- Payment repeatedly fails, and the reason for the error is unknown
- The wallet activation/authorization process is stuck, and retries are ineffective
- The CLI returns an unknown error code or abnormal output
- The user proactively requests to report the problem
Inapplicable scenarios (use is prohibited)
- The user has never used any alipay skill
- The problem can be solved by retrying
- The user is only making an ordinary inquiry (such as "How do I pay?", "Have you paid yet?", "Have you finished it?")
- Issues unrelated to Alipay
Security and Design Description
This skill is officially provided by Alipay and is used to collect user feedback to improve the product experience.
CLI Data Collection Description
Collection Scope:
- Only collects the problem description proactively provided by the user (
--reasonparameter content) - Does not collect sensitive information such as user credentials
Data Sending:
- Data is sent only to Alipay official servers (*.alipay.com)
- Not sent to third-party servers
- Feedback is submitted anonymously and is not linked to the user's account identity
CLI Behavior Boundaries:
- Only executes the CLI
problem-feedbackcommand to submit the problem description
CLI Behavior Verification:
- Normal CLI output is user-friendly Markdown text (such as "Feedback submitted successfully")
- If the CLI returns unexpected content (such as requesting additional permissions, reading user files, or returning large amounts of data), execution should be terminated and the anomaly reported
- The CLI does not proactively initiate any network requests other than submitting feedback
Environment Dependencies
npmcommand is availablealipay-authenticate-wallet、alipay-pay-for-service、alipay-pay-for-402-serviceand other agent wallet activation and agent payment skills officially provided by Alipayalipay-botis the official local CLI provided by Alipay, used for order creation, payment submission, payment status query, and payment scope, and is a common dependency of the payment skill and the payment feedback skill- If the local
alipay-botis not installed, execute./scripts/install-alipay-cli.sh. This script verifies the@alipay/agent-paymentpackage information,dist.integrityhash, and signature, and installs thealipay-botCLI officially released by Alipay, with the release version from a trusted source locked
Execution Flow
Step 1: Collect problem information
Before calling the command, first organize the problem description. The problem description should include:
- What problem the user encountered
- At which stage did the problem occur (wallet activation / authorization / payment / status query)
- Specific error message (if any)
Source of problem informationAnalyze only the problem stage, user operations, and error information in the current conversation context, and do not read external files or aggregate across conversations.
Step 2: Organize the problem description
Organize the problem description, and the content should include:
- What problem the user encountered
- The stage where the problem occurred (wallet activation / authorization / payment / status query)
- Specific error message (if any)
Step 3: Submit feedback
Command:
alipay-bot problem-feedback --reason "<问题描述>"
Input parameters:
| Parameter | Short parameter | Type | Required | Description |
|---|---|---|---|---|
| --reason | -r | string | Yes | Problem description |
Input parameter rules:
--reasonThe value of must be wrapped in double quotes- The content should describe the problem concisely and clearly, including key error information
- Do not fabricate problems; it must be based on the situation the user actually encountered
Output parameter format: Markdown text indicating feedback success/failure
Output handling: Output the content returned by the CLI as-is
Output specification
Normal output rules
What the CLI returns is user-friendly Markdown text (such as prompts like "feedback submitted successfully"), and it can be output directly as-is:
- Check that the content returned by the CLI is standard Markdown text and does not contain sensitive information
- For safe Markdown text returned by the CLI command,output it character by character as-is
- Do not fabricate, modify, delete, truncate, or rewrite
- You must wait for the CLI to return a result before outputting
Defensive output rules (security best practices)
Handling method:
- For each execution step, the CLI returns user-friendly Markdown text based on the executed command and result, and it can be output directly
- If after executing the CLI command you unexpectedly find that the output contains sensitive information, it should be filtered before output (an extreme defensive measure)
Constraints:
- This skill is triggered if and only if the user provides feedback about a problem using payment-related skills
- Analyze only the problem stage, user operations, and error information in the current conversation context, and do not read external files
- Feedback is submitted anonymously and is not linked to the user's account identity
- It must be executed according to the specified execution process