Channels Guide logo

Channels Guide

Community
wentorai
Channels Guide

Configuration, troubleshooting, and operational guide for Research-Claw IM channels (Telegram, Discord, WeChat, Feishu, Slack, WhatsApp). Covers dashboard UI flow, bot token setup, QR login, config.patch restart behavior, enable/disable/delete, common errors (401, ABI, plugins.allow), and diagnostic procedures.

Overview

Publisherwentorai
RepositoryResearch-Claw
Skill nameChannels Guide
Stars
859
Forks
117
Bundled files
Instructions only
Links
  • Markdown instructions

    A SKILL.md file the model loads on demand, so it only costs tokens when a request actually matches.

  • Works with any LLM

    AI skills are plain Markdown, not provider-specific code, so this works with GPT, Claude, Gemini, Grok, or a local model.

  • Self-contained

    Everything the model needs lives in the instructions — no extra files to sync.

  • Open source

    Published by wentorai on GitHub. Read the source before you install it.

Installation

Install the Channels Guide AI skill in TypingMind to use it with any LLM, or drop it into another agent that reads SKILL.md.

1

Install in TypingMind

TypingMind installs a skill straight from its GitHub folder — it reads SKILL.md, bundles the resource files, and stores the result locally.

  1. Open the app and go to Plugins → Skills.
  2. Choose "Install from GitHub".
  3. Paste the skill folder URL below and confirm.
    https://github.com/wentorai/Research-Claw/tree/main/skills/channels-guide
  4. Enable the skill in any chat where you want it available.
Plugins → Skills → Add skill → From GitHub URL, then paste the folder URL and press Continue.
2

Install in another agent

Any agent that reads the Agent Skills format can use this skill — copy the folder into that agent's skills directory.

Claude Code — .claude/skills
git clone --depth 1 https://github.com/wentorai/Research-Claw.git /tmp/Research-Claw
mkdir -p .claude/skills
cp -r /tmp/Research-Claw/skills/channels-guide .claude/skills/wentorai-channels-guide
Restart Claude Code after copying so it picks up the new skill.

Use it in TypingMind

Enable Channels Guide in any TypingMind chat and the model takes it from there. Its name and description sit in the system prompt, and the moment a request matches, the model loads the full instructions itself — you never invoke it by hand, and it costs no tokens until it is actually used.

The model loads Channels Guide on its own as soon as a request matches it.

Works with any AI model

AI skills are plain Markdown instructions rather than provider-specific code, so Channels Guide is not tied to the model it was written for. Install it once in TypingMind and use it with GPT-5, Claude, Gemini, Grok, DeepSeek, Mistral, Llama, or a local model you run yourself — all on your own API keys.

  • Loaded only when it is needed

    The system prompt carries just the name and description. The instructions are fetched on the first matching request, so an idle skill costs nothing.

  • Switch models mid-chat

    Because the skill is instructions rather than code, changing model does not break it — the next model reads the same SKILL.md.

Skill instructions

This is the SKILL.md content the model loads. Read it before installing — a skill is instructions your model will follow.

IM Channel Configuration

RC can receive/reply via Telegram, Discord, WeChat (微信), Feishu (飞书), QQ, Slack, WhatsApp. Channels are OC infrastructure — RC reuses them fully.

Dashboard Channel Management UI

The dashboard has a complete channel management UI in the left sidebar.

Location: Left sidebar → Extensions panel (🔗 chain-link icon) → Channels tab.

What users can do in the Channels tab:

  • View all channels with live status (connected / configured / error / disabled)
  • WeChat / WhatsApp: Click the Connect button → QR code modal opens → scan with phone
  • Enable / Disable any channel via a toggle switch (triggers config.patch → gateway restart)
  • Delete a channel via the red trash button (with confirmation modal)
  • Add Channel button at the bottom (sends a chat message asking you for help)
  • Ask Agent to Fix / Configure button on error/unconfigured channels (sends chat message)

How to guide users:

  • For WeChat / WhatsApp QR login: Direct users to the Channels tab: "请打开左侧扩展面板 → 通道标签 → 点击微信旁边的 Connect 按钮,在弹出窗口中扫码。" Do NOT generate QR codes in the chat — the dashboard QR modal handles it.
  • For bot-token channels (Telegram, Discord, etc.): Help via chat — guide them to get a token, then use config.patch to write it. Or they can click "Ask Agent to Configure".
  • The weixin_login / whatsapp_login agent tools are fallbacks for IM-only contexts (when the user is chatting from Telegram/Discord and has no dashboard access).

Connection Protocols

Bot-Token Type (Telegram / Discord / Feishu / QQ / Slack)

  1. Guide user to create a bot on the platform and obtain a token.
    • Telegram: Talk to @BotFather → /newbot → copy token.
    • Discord: Discord Developer Portal → Bot section → copy token.
    • Feishu: Open platform → create bot → App ID + App Secret.
  2. Write config via config.patch:
    • Telegram: { channels: { telegram: { botToken: "<TOKEN>", enabled: true } } }
    • Discord: { channels: { discord: { token: "<TOKEN>", enabled: true } } }
    • Feishu: { channels: { feishu: { appId: "...", appSecret: "...", enabled: true } } }
    • Others: see platform OC docs for field names.
  3. After config.patch: Gateway auto-restarts via SIGUSR1 (takes ~3–5s). The dashboard reconnects automatically. No manual restart needed.
  4. Telegram: user must send "/start" in the bot chat to receive replies.
  5. commands.native must be false (530+ tools exceed IM menu limits). sync-global-config.cjs auto-fixes this on startup.

WeChat (微信) — QR Scan

Prerequisite: openclaw-weixin plugin installed and in plugins.allow.

Primary path (dashboard): User clicks Connect in the Channels tab → QR modal → scan. The dashboard handles web.login.start → web.login.wait automatically, then triggers config.patch to restart gateway and pick up new credentials.

Fallback path (IM-only / no dashboard): Use the weixin_login agent tool:

  1. Call weixin_login { action: "start" }.
    • Returns QR code URL in markdown: ![weixin-qr](https://...).
    • Display the image to the user and prompt them to scan with WeChat.
  2. Call weixin_login { action: "wait", timeoutMs: 120000 }.
    • Blocks until the user scans or timeout.
    • On success, plugin auto-saves credentials; gateway auto-starts channel.
  3. WeChat cannot send proactive messages — replies only (contextToken mechanism).

WhatsApp — QR Scan

Primary path: Same as WeChat — user clicks Connect in the Channels tab.

Fallback path (IM-only): Use the whatsapp_login agent tool:

  1. Call whatsapp_login { action: "start" }.
    • Returns QR code as base64 data URL in markdown: ![qr](data:image/png;base64,...).
  2. Call whatsapp_login { action: "wait", timeoutMs: 120000 }.
    • Blocks until the user scans or timeout.

Config & Restart Behavior

Understanding this is critical for troubleshooting:

  1. config.patch → gateway receives SIGUSR1 → drains active tasks (up to 90s) → restarts → channels reload from new config. Dashboard auto-reconnects.
  2. Enable/Disable switch in dashboard → calls config.patch with { channels: { <id>: { enabled: true/false } } } → restart.
  3. Delete button → calls config.patch with { channels: { <id>: null } } → restart.
  4. If a channel has an invalid token (e.g., placeholder <YOUR_TOKEN>), the channel will crash on startup with 401 Unauthorized and enter an auto-restart loop (exponential backoff: 5s → 10s → 20s → ... up to 10 attempts). This is normal; it stops after 10 failures. Fix: provide correct token or disable the channel.
  5. Drain can be slow (up to 90s) if there are active agent tasks or embedded runs. This is by design — gateway waits for work to complete before restarting.

In-Channel Behavior

  • All RC tools (library, tasks, workspace, monitor) are fully available.
  • Keep replies under 2000 characters (IM message limit).
  • Do not use Markdown tables (most IM clients do not render them).
  • approval_card degrades to text: "需要审批: xxx. 回复 yes/no".
  • Media requires absolute paths via the media parameter.
  • Peer ID formats differ: WeChat xxx@im.wechat, Telegram numeric IDs.

Troubleshooting

Common Errors

SymptomCauseFix
401: Unauthorized loopInvalid or placeholder bot tokenProvide real token via config.patch or disable channel
"not configured"No credentials found for channelFor bot-token: set token in config. For WeChat: re-scan QR
"Provider not running"Channel configured but crashedCheck logs for error. Common: token expired, network issue
Channel not visiblePlugin not in plugins.allow or plugins.load.pathsAdd to both in config. See 4-point sync below
better-sqlite3 crashNative module ABI mismatchRebuild under conda Node 22 (gateway's Node version)
WeChat "not configured" after QROld default account has no tokenRe-scan QR. Dashboard uses best account for status

Diagnostic Steps

When a channel shows Error or Not Configured:

  1. Check the token: Read config via config.get or ask user. Is it a real token or a placeholder like <YOUR_TOKEN>? Placeholder = 401 errors.
  2. Check plugins.allow: The channel plugin ID must be in plugins.allow array. Example: openclaw-weixin for WeChat, telegram for Telegram (OC built-in).
  3. Check credentials path (WeChat only): Stored at ~/.openclaw/openclaw-weixin/accounts/<accountId>.json. If missing, user needs to re-scan QR.
  4. Check gateway logs: Look for [telegram], [openclaw-weixin] etc. in terminal output. Common patterns:
    • 401: Unauthorized → invalid token
    • ENOTFOUND / ETIMEDOUT → network/VPN issue
    • channel exited → channel crashed, will auto-restart
  5. Restart gateway if needed: gateway.restart (requires approval_card). Or user can Ctrl+C and re-run run.sh.

4-Point Plugin Sync (for new channel plugins)

When adding a new channel plugin, it must be registered in 4 places:

  1. config/openclaw.json → plugins.allow array
  2. config/openclaw.json → plugins.load.paths array
  3. scripts/ensure-config.cjs → REQUIRED_ALLOW array
  4. dashboard/src/utils/config-patch.ts → RC_CONFIG_DEFAULTS.plugins.allow Plus the runtime global ~/.openclaw/openclaw.json.

Frequently asked questions

What does the Channels Guide AI skill do?

Configuration, troubleshooting, and operational guide for Research-Claw IM channels (Telegram, Discord, WeChat, Feishu, Slack, WhatsApp). Covers dashboard UI flow, bot token setup, QR login, config.patch restart behavior, enable/disable/delete, common errors (401, ABI, plugins.allow), and diagnostic procedures.

Why use Channels Guide on TypingMind?

Because you install it once and use it with any model. Channels Guide is plain Markdown rather than provider-specific code, so the same skill runs on GPT-5, Claude, Gemini, Grok, or a local model — and you can switch model mid-chat without it breaking. TypingMind runs on your own API keys, so you pay providers directly instead of a per-seat subscription, and your skills and chats stay in your own storage.

How do I install Channels Guide in TypingMind?

Open Plugins → Skills → Install from GitHub in TypingMind and paste https://github.com/wentorai/Research-Claw/tree/main/skills/channels-guide. TypingMind reads its SKILL.md and installs it as a skill you can enable per chat.

Which AI models can use Channels Guide?

Any model you connect in TypingMind. AI skills are plain Markdown instructions rather than provider-specific code, so GPT, Claude, Gemini, Grok, and local models can all load this skill when a request matches it.

How many AI models can I use with Channels Guide?

As many as you like. As long as a model supports skills, you can use Channels Guide with it — GPT, Claude, Gemini, Grok, DeepSeek, Mistral, Llama and more — all on TypingMind with your own API keys.

Is the Channels Guide AI skill free?

It is published on GitHub by wentorai. Check the repository for licensing terms. You only pay your own AI provider for the tokens you use.

What are AI skills?

An AI skill is a reusable instruction bundle that teaches an AI model how to do one specific task. It follows the open Agent Skills format: a SKILL.md file with a name and description, plus any scripts, templates or reference files the model may need. The model reads the instructions only when your request matches the skill, so an installed skill costs nothing until it is used.

How are AI skills different from plugins or MCP servers?

A plugin or MCP server gives a model new tools to call — code that runs somewhere and returns a result. An AI skill gives the model knowledge and process instead: how to approach a task, which steps to follow, what good output looks like. Skills are plain Markdown, so they need no server, no API key and no runtime, and they work with any model.

View all

Set up your own AI workspace now

Get notified about new features and future giveaways by subscribing to our newsletter 👇