Access split
DMs vs shared spaces
A working DM path does not prove the bot is ready in channels, rooms, or groups.
Mention gating
A healthy bot can still look quiet when shared-space replies require an explicit mention.
OpenClaw channels
Cross-channel setup patterns
Use this page when you want the reusable setup mental model before you dive into platform-specific details.
Core idea
Most channel setup is the same problem in different clothes.
The repeated pattern
What setup work usually repeats everywhere
- get working credentials
- make sure the runtime can resolve them
- decide direct-message policy
- decide shared-space policy
- decide mention behavior
- test each path separately
Safe rollout order
Use a setup ladder that is easy to debug
A good default order is obtain credentials, store them in a runtime-readable way, enable the channel in config, choose DM policy, choose shared-space policy, apply and verify health, test DMs, test shared spaces, and only then widen access intentionally.
What repeats across channels
Shared concepts behind platform-specific details
Visibility split
Identity split
Different ID formats, same policy pattern
Discord, Telegram, WhatsApp, Slack, and Matrix use different IDs, but the allow/deny model repeats.
Next route
Where this lane should grow next
Policy explainer
Pairing, allowlists, and mention gating
Separate the three controls cleanly.
Diagnosis page
Connected but not replying
Handle false-success channel states directly.
Platform pages
Discord / Telegram / WhatsApp
Add concrete setup flows once the shared mental model is in place.