WxPusher Integration

    1

    Integrate and troubleshoot WxPusher notifications with SPT JSON POST, preserving slashes and URLs; covers UID/Topic, response checks, credentials, and receiving-channel limits.

    Free

    1 useSecurity scanned

    Works with the AI Agents you already use

    Claude CodeClaude CodeCursorCursorCodex CLICodex CLIMuseMuseDotsDots+21 more

    WxPusher Integration

    Example session with this skill installed

    WxPusher is already selected, and I have authorized notifications to my existing SPT.
    Our current script uses SPT GET. Prepare the request for this text, preserving it exactly:

    【需要处理】请执行 /context,然后查看 https://example.invalid/a/b?next=1#frag,百分号 100%。
    保留原文。

    Use only the placeholder SPT_example and do not send a real message. Then interpret a
    mock HTTP 200 response with top-level code 1001, and a mock response with top-level code
    1000 but one successful and one failed target.

    • Read your context and instructions
    • Compiled the wxpusher integration

    Request choice

    Use the existing SPT with JSON POST. The current GET path limitation affects the slashes
    in /context and the URL. A single-recipient request needs no appToken or UID, and this
    already authorized use does not require another authorization prompt.

    POST https://wxpusher.zjiecode.com/api/send/message/simple-push
    Content-Type: application/json
    
    {
      "spt": "SPT_example",
      "content": "【需要处理】请执行 /context,然后查看 https://example.invalid/a/b?next=1#frag,百分号 100%。\n保留原文。",
      "contentType": 1
    }
    

    Serialize this object as JSON and encode the request body as UTF-8. Decoding the body
    returns exactly the original text, including Chinese, slashes, the URL, percent sign,
    and newline. Do not replace slashes or URL-encode the content. The \n escape is JSON's
    representation of the original newline.

    Mock-response interpretation

    Supplied responseOverall outcome
    HTTP 200, top-level code 1001Business failure; do not report API acceptance or delivery success.
    HTTP 200, code 1000, targets code 1000 and 1001Partial failure; do not report that every target was accepted.
    HTTP 200, code 1000, every returned target code 1000API acceptance only; receiving or reading is unconfirmed.

    Inspect sanitized business errors and recipient configuration. Follow the project's
    bounded policy for eligible transient failures and account for duplicate delivery before
    retrying. Authentication, validation, and permission failures are not reasons for blind
    retries. Do not call retired send-status endpoints or resend because of their retirement
    notice. GET remains available for suitable simple text; the current slash limitation is
    separate from correct URL encoding.

    This is an illustrative analysis using the independently checked request scenario and
    controlled response fixtures. No WxPusher API request or real delivery was performed.

    Connects securely to your tools. The creator never sees your data.

    About this skill

    Author version: 1.0.5 · License: MIT · Primarily in Simplified Chinese.

    WxPusher Integration is an Agent Skill for adding and troubleshooting WxPusher notifications in scripts, backend services, scheduled tasks, and CI. Use it when the user has explicitly chosen WxPusher. Generic UID or Topic terms, or an unspecified request for WeChat notifications, do not independently select this provider.

    Version 1.0.5 defaults existing SPT workflows to JSON POST at https://wxpusher.zjiecode.com/api/send/message/simple-push. A single-recipient request uses the existing spt, original content, and contentType. It preserves /context, https://..., Chinese text, newlines, and percent signs through JSON serialization and UTF-8 encoding, without URL percent-encoding the content or switching to appToken/UID.

    SPT GET remains a limited convenience endpoint. Content containing slashes, which become %2F in the path, should currently use POST. Do not work around this limitation with double encoding, character replacement, or repeated requests. The current limitation is recorded separately from the public documentation: it is supported by a prior invalid-token test summary supplied by the user and the user's account of an official staff reply. The reply has no public link or original support transcript; public documentation still lists GET and does not explicitly describe this limitation.

    The skill also covers standard JSON push, UID/Topic targeting, content formats and limits, credential handling, response classification, bounded retries, and receiving-channel troubleshooting. HTTP 200 is insufficient: check the top-level business code and every returned target result. Acceptance means a task was created, not that the recipient read or received it. The two legacy send-status query endpoints are now documented as disabled; stop polling them and do not resend because of their retirement notice.

    WeChat ClawBot/iLink text notifications require binding, enabling, and activation. Existing channel limits still apply, and that channel currently does not support upstream messages. This is not arbitrary friend or group messaging.

    The package consists of 8 Markdown/YAML files, including SKILL.md, references, optional UI metadata, and the complete MIT license. It contains no executable adapter or bundled test suite and has no Python, Node.js, or PowerShell runtime dependency. A compatible agent host provides file, HTTP, shell, and development tools for the authorized project task. Users provide their own WxPusher credentials through their approved local configuration or secret store. Existing, applicable long-term authorization is preserved.

    Public SPT documentation, send responses, and retired status queries were rechecked on 2026-11-12; other API/SDK facts retain their previous verification dates. Windows checks covered structure, 18 local references, YAML/JSON examples, credential patterns, whitespace, file integrity, two independent request scenarios, and six controlled response fixtures. UTF-8 JSON round trips preserve the original message, and the local installed copy matches the source. No real WxPusher delivery test was performed; these checks do not establish end-to-end delivery or real-device compatibility for every host.

    Independently maintained; this is not an official WxPusher project.

    Source: https://github.com/hunxuankai/wxpusher-integration/tree/634107ef4e5c27ad817356c6c1bc1c599f86f075/skills/wxpusher-integration

    WxPusher must be explicitly selected. Live sends need the user's own configuration;
    placeholders suffice for implementation and mock tests. Current SPT GET content with
    encoded slashes uses POST instead; GET is not entirely disabled. WeChat ClawBot requires
    binding, enabling, and activation, and allows up to 10 text notifications within 24 hours
    after activation, with reactivation when exhausted; it does not provide arbitrary friend
    or group messaging or upstream messages. API acceptance does not prove receipt or reading.
    Two legacy send-status queries are retired. Guides are primarily in Simplified Chinese.
    SPT/send-response/retired-query public sources were rechecked on 2026-11-12; other API/SDK
    facts retain their original dates. Validation was offline and does not establish
    end-to-end delivery or real-device compatibility for every host.

    How to use

    1. 1

      Connect your agent to the Agensi MCP

      Copy the prompt and paste it into Claude, ChatGPT, Cursor or any agent that supports MCP. It sets up the connection for you. Manual setup

    2. 2

      Sign in to Agensi

      Your agent shows a sign-in link. Sign in and click Allow.

    3. 3

      Ask for what you need

      Your agent loads this skill and uses it on your task. It always asks before any payment.

    Reviews

    No reviews yet

    Be one of the first to try it. Every listed skill goes through the trust checks below.

    Security scanned

    Checked by our 8-point scan before listing

    1 use

    Times used to date

    Free forever

    No account required to browse

    Trust & safety

    Security scan

    Security scan passed 4 days ago

    • Free with an Agensi account

    Requires

    Git

    Permissions

    Read Files
    Write Files
    Network Access
    Environment Variables

    Reads and edits only the project files needed for the authorized integration; uses existing shell/test tools and an HTTP client. WxPusher credentials come from user-approved environment variables or secret storage. Network access supports official documentation and authorized API calls; development validation defaults to mocked responses. Existing applicable user authorization is respected.

    Listed1 month ago
    Updated4 days ago

    What's inside

    Frequently Asked Questions

    Author version 1.0.5 includes 8 Markdown/YAML files: SKILL.md, LICENSE.md, five reference guides, and optional agents/openai.yaml metadata. It provides integration instructions and examples, without a bundled executable adapter or test suite. Existing SPT workflows default to UTF-8 JSON POST while preserving the original content. Agensi package revisions are numbered separately from the author version.

    It covers WxPusher clients and WeChat ClawBot (iLink) text notifications. WeChat requires binding, enabling, and activation, and is subject to channel limits. This is not an API for arbitrary WeChat contacts or group chats.

    The complete folder follows the open SKILL.md format and does not require a particular host. Guides are primarily in Simplified Chinese. Integration tasks use the agent’s existing project and HTTP tools; compatibility with every agent has not been individually runtime-tested.

    No. Implementation and mock testing can use named environment variables and placeholders. Live delivery requires your own appToken plus the appropriate UID or Topic, or an SPT for the selected SPT workflow. Credentials are not included in this package.

    Check HTTP status, JSON, the top-level business code, and every returned target code. HTTP 200 with a business failure is not success; mixed target results are partial failure. Acceptance does not prove receipt or reading. Use bounded retries only for eligible transient failures and consider duplicates. Do not call retired status queries or resend because of their retirement notice.

    This is a free, MIT-licensed, independently maintained skill, not an official WxPusher project. SPT requests, send responses, and retired status queries were rechecked against public documentation on 2026-11-12; other API/SDK facts retain their original verification dates. Offline validation does not prove real delivery. Existing applicable user authorization governs real sends.

    New in APIs & Backend