Nodes/comfyui_LLM_party/send to wechat official
ComfyUI Node

send to wechat official

Publish your agent's output straight to a WeChat account

By heshengtao·Created 2 years ago·Updated 7 days ago· 2,321
send to wechat official
    • log
    appidappid
    secretsecret
    titletitle
    authorauthor
    contentHTML content
    is_enabletrue

    LLM Party isn't just about generating text - a chunk of the pack is dedicated to shipping whatever your agent produced out to the platforms people actually read things on. send_to_wechat_official is the WeChat leg of that: it posts content directly to a WeChat Official Account (公众号), the publishing accounts businesses and creators in China use to reach subscribers. If your workflow's whole point is "LLM writes it, node publishes it," this is the last step.

    It's a straightforward API call under the hood: authenticate with your account's credentials, send title/author/content, get a result back. Nothing fancy - the value is that it's already wired into a graph where an LLM node upstream can be the one writing title and content in the first place.

    Inputs and outputs that matter

    • appid and secret - your WeChat Official Account's API credentials, from the account's backend at mp.weixin.qq.com. Not something you invent; you need a real registered account to get these.
    • title and author - plain text, straightforward.
    • content - the node's default value is literally "HTML content", which tells you what it expects: HTML markup, not plain text. If you've got an LLM writing rich copy, or you've run something through this same pack's svg2html node, that's the shape you're feeding in here.
    • is_enable - the kill switch. Almost every node in this pack has one; flip it off to disable without unplugging wires.

    The output is log - a string, presumably whatever status or response the WeChat API handed back, useful for confirming the publish actually went through (or seeing why it didn't) without stopping the rest of your graph.

    Installing it

    Standard LLM Party install: search comfyui_LLM_party in ComfyUI Manager, or git clone https://github.com/heshengtao/comfyui_LLM_party into custom_nodes and pip install -r requirements.txt inside your ComfyUI environment, then restart. This node itself has no model or GPU dependency - it's a plain HTTP call - so it's one of the lighter reasons to have this heavyweight pack installed.

    Common issues

    This is China-platform-specific, so the first "issue" most people will hit is simply not having a WeChat Official Account to point it at - you can't test this node meaningfully without one. Assuming you do: WeChat's Official Account API whitelists calling IPs from the account's backend settings. If you're running ComfyUI on a machine whose outbound IP isn't the one you whitelisted - or worse, an environment where the IP isn't fixed at all - the call will fail with an authentication-shaped error that has nothing to do with your appid/secret actually being wrong. Add the right IP in the account backend before you spend time debugging the node itself.

    The other common trap is the account tier. A lot of the more useful publish-style WeChat APIs are gated to verified/service accounts rather than basic personal or unverified subscription accounts - if calls keep failing for no obvious reason, check what your account type is actually allowed to do before assuming it's a credentials problem. And because content needs to be real HTML, a plain-text string dropped straight from an LLM node will "work" (no error) but render as an unstyled wall of text on the published side - wrap it properly, or run it through an HTML-producing step first.

    Category大模型派对(llm_party)/APP链接(app link)

    Inputs (6)

    NameTypeDefaultDescription
    appidSTRINGappid
    secretSTRINGsecret
    titleSTRINGtitle
    authorSTRINGauthor
    contentSTRINGHTML content
    is_enableBOOLEANtrue

    Outputs (1)

    NameTypeDescription
    logSTRING