A practical walkthrough of Claude Code channels: how an MCP-server channel pushes messages, alerts, and webhooks into a running session, how to install and enable Telegram, Discord, or iMessage, sender allowlists, and enterprise controls.
Open the source and read safety notes before installing.
Source-backed facts for citing this resource, derived directly from the registry — also available as plain text for AI assistants.
Decision playbook
Signals are present but mixed. Use the checklist below to confirm the source and operational safety for your environment.
0
63
—
No baseline selected
No major trust-signal divergence detected in the current selection.
Confirm ownership and provenance before trusting install instructions.
Source link availableRequired
Open the canonical repository and verify ownership.
Source provenance statusRequired
Marked as source-backed.
Metadata reviewed
No reviewed flag detected in metadata.
Validate risk disclosures before installation or API wiring.
Safety notes presentRequired
Review the listed safety guidance before running commands.
Privacy notes presentRequired
Review data handling notes before connecting accounts or secrets.
Trust level risk gateRequired
Trust level does not block evaluation.
Check package metadata and artifact integrity signals.
Install payload available
Install or copy payload is available for review.
Package verification flag
No package verification flag provided.
Checksum metadata
No checksum provided for downloaded artifact.
Use compare context to validate trade-offs before adoption.
Compare tray has multiple entries
Add at least one more entry to compare trust differences.
Baseline comparison available
No baseline peer selected yet.
Diverging trust signals identified
No major trust-signal divergence found.
Setup at a glance
Copy-ready — paste the snippet to get started.
Adoption plan
Current risk score 24/100. Use staged verification before broader rollout.
Validate source and review signals before any execution.
Confirm source provenanceRequired
Source URL/provenance metadata is present.
Confirm metadata review state
No review metadata found; increase manual validation.
Verify install payload
Install/config payload exists and can be inspected.
Confirm safety, privacy, and package integrity signals.
Review safety notesRequired
Safety notes are present.
Review privacy notesRequired
Privacy notes are present.
Verify package integrity metadata
No package verification/checksum metadata.
Adopt in controlled steps based on the selected plan.
Run in isolated sandbox firstRequired
Use a constrained sandbox and observe behavior across multiple tasks.
Roll out graduallyRequired
Roll out to a small cohort before wider usage.
Set monitoring and fallback
Define rollback path and monitor errors after adoption.
Evidence readiness
Missing required evidence: Metadata review. Risk score 31.
Source repository/provenance is listed.
Required in this preset
Review metadata is missing.
Required in this preset
Safety notes are present.
Required in this preset
Privacy notes are present.
Optional in this preset
Package integrity metadata is missing.
Optional in this preset
Install payload is available.
Required in this preset
Required gaps: Metadata review
Decision timeline
Blocking gaps: Check metadata review status. Risk 28.
triage
Source/provenance metadata is available.
triage
Review metadata is missing.
verify
Safety notes are available.
verify
Privacy notes are available.
verify
Package integrity metadata is missing.
rollout
Install payload is available.
Blockers: Check metadata review status
Prerequisite readiness
3 prerequisites to line up before setup. Have accounts and credentials ready first.
Safety & privacy surface
3 safety and 3 privacy notes across 4 risk areas. Review closely: credentials & tokens, permissions & scopes, third-party handling.
## Overview
A channel is an MCP server that pushes events into your running Claude Code
session, so Claude can react to things that happen while you are away from the
terminal. Channels can be two-way: Claude reads the event and replies through the
same channel. Events only arrive while the session is open, so for an always-on
setup you run Claude in a background or persistent terminal. Channels are a
research preview and require Anthropic authentication through claude.ai or a
Console API key.
## Install and enable a channel
Channel plugins require Bun. Using Telegram as an example:
1. Create a bot with BotFather and copy the token.
2. Install the plugin: `/plugin install telegram@claude-plugins-official` (run
`/reload-plugins` afterward).
3. Configure the token: `/telegram:configure <token>` (saved to
`~/.claude/channels/telegram/.env`).
4. Restart with the channel enabled:
```bash
claude --channels plugin:telegram@claude-plugins-official
```
5. Message your bot to get a pairing code, then run
`/telegram:access pair <code>` and lock it down with
`/telegram:access policy allowlist`.
Discord and iMessage follow the same pattern. Try the `fakechat` localhost demo
first to learn the flow with nothing to authenticate.
## Webhook receiver
Because a channel is an MCP server, you can build one that receives webhooks from
CI, an error tracker, or a deploy pipeline, so the event arrives in the session
where Claude already has your files open and remembers what you were debugging.
See the Channels reference for building your own.
## Security
- Every channel keeps a sender allowlist; only paired or explicitly allowed IDs
can push messages, and everyone else is silently dropped.
- Being listed in `.mcp.json` is not enough; a server must also be named in
`--channels` for the session.
- If permission relay is enabled, anyone who can reply through the channel can
approve or deny tool use, so allowlist only senders you trust with that.
## Enterprise controls
Admins govern availability with managed settings:
- `channelsEnabled` is the master switch (blocked by default on Team and
Enterprise; allowed by default on Console unless managed settings are
deployed).
- `allowedChannelPlugins` replaces the default allowlist with a specific set of
approved plugins and marketplaces.
In all cases, no channel runs until a user opts it in with `--channels`.
## How channels compare
Channels push events into your already-running local session. That differs from
Claude Code on the web (fresh cloud session), Claude in Slack (spawns a web
session from a mention), a standard MCP server (queried on demand, nothing
pushed), and Remote Control (you drive the local session from your phone).
## Source
- Push events into a running session with channels: https://code.claude.com/docs/en/channelsA channel is an MCP server that pushes events into your running Claude Code session, so Claude can react to things that happen while you are away from the terminal. Channels can be two-way: Claude reads the event and replies through the same channel. Events only arrive while the session is open, so for an always-on setup you run Claude in a background or persistent terminal. Channels are a research preview and require Anthropic authentication through claude.ai or a Console API key.
Channel plugins require Bun. Using Telegram as an example:
/plugin install telegram@claude-plugins-official (run
/reload-plugins afterward)./telegram:configure <token> (saved to
~/.claude/channels/telegram/.env).claude --channels plugin:telegram@claude-plugins-official
/telegram:access pair <code> and lock it down with
/telegram:access policy allowlist.Discord and iMessage follow the same pattern. Try the fakechat localhost demo
first to learn the flow with nothing to authenticate.
Because a channel is an MCP server, you can build one that receives webhooks from CI, an error tracker, or a deploy pipeline, so the event arrives in the session where Claude already has your files open and remembers what you were debugging. See the Channels reference for building your own.
.mcp.json is not enough; a server must also be named in
--channels for the session.Admins govern availability with managed settings:
channelsEnabled is the master switch (blocked by default on Team and
Enterprise; allowed by default on Console unless managed settings are
deployed).allowedChannelPlugins replaces the default allowlist with a specific set of
approved plugins and marketplaces.In all cases, no channel runs until a user opts it in with --channels.
Channels push events into your already-running local session. That differs from Claude Code on the web (fresh cloud session), Claude in Slack (spawns a web session from a mention), a standard MCP server (queried on demand, nothing pushed), and Remote Control (you drive the local session from your phone).
Show that Channels and Webhooks into Running Claude Code Sessions is listed on HeyClaude. Paste this Markdown into your README — it renders the badge and links back to this page.
[](https://heyclau.de/entry/guides/channels-and-webhooks-into-running-claude-code-sessions)Channels and Webhooks into Running Claude Code Sessions side by side with 3 alternatives on trust, install, platform support, and disclosed safety notes — all from reviewed registry metadata.
2 trust signals differ across this comparison (Source provenance, Submitter).
| Field | A practical walkthrough of Claude Code channels: how an MCP-server channel pushes messages, alerts, and webhooks into a running session, how to install and enable Telegram, Discord, or iMessage, sender allowlists, and enterprise controls. Open dossier | Use Claude Code agent view (claude agents) to dispatch, monitor, peek, attach, and organize many background sessions from one terminal with state icons, PR labels, filters, and shell commands for fleet-style multitasking. Open dossier | Guide to /compact, /memory, and CLAUDE.md hygiene for long Claude Code sessions: when to compact, what to store in memory, and avoiding stale context. Open dossier | A practical guide for keeping Claude Code sessions focused during long coding work with scoped prompts, checkpoints, durable memory boundaries, source refreshes, and privacy-safe handoffs. Open dossier |
|---|---|---|---|---|
| Next steps | ||||
| Trust | ||||
| Review status | Not reviewed | Not reviewed | Not reviewed | Not reviewed |
| Package trust | Package not verified | Package not verified | Package not verified | Package not verified |
| Source provenanceDiffers | Source-backed | Submission linkedSource submission | Submission linkedSource submission | Source-backed |
| SubmitterDiffers | JPette1783 | kiannidev | kiannidev | MkDev11 |
| Install risk | Review first | Review first | Review first | Review first |
| Notes | Safety ✓ Privacy ✓ | Safety ✓ Privacy ✓ | Safety ✓ Privacy ✓ | Safety ✓ Privacy ✓ |
| Brand | — | — | — | — |
| Category | guides | guides | guides | guides |
| Source | Source-backed | Source-backed | Source-backed | Source-backed |
| Author | JPette1783 | kiannidev | kiannidev | MkDev11 |
| Added | 2026-06-05 | 2026-06-16 | 2026-06-14 | 2026-06-04 |
| Platforms | ||||
| Harness | ||||
| Source repo | — | — | — | — |
| Safety notes | ✓Channel input is untrusted external content; treat incoming messages as data, not as instructions that should widen permissions. Events only arrive while the session is open; for unattended use, --dangerously-skip-permissions bypasses prompts and should be used only in trusted environments. Each channel keeps a sender allowlist; only paired/allowlisted IDs can push messages, and anyone who can reply can also approve permission relays if enabled, so allowlist only trusted senders. | ✓Background sessions run with the permission mode shown in the agent view header; review bypassPermissions and auto modes before dispatching unattended work. Deleting a session from agent view can remove Claude-created worktrees and uncommitted changes—commit or push first. Each dispatched session is a full Claude Code process; do not parallelize destructive tasks on the same checkout without isolation. | ✓Compaction summarizes conversation history; verify summaries before relying on them for security-sensitive or release-critical decisions. Do not store secrets, credentials, or customer identifiers in `/memory` or CLAUDE.md. After compaction, re-validate repository state, test results, and open PR status before continuing destructive edits. | ✓Do not paste credentials, production secrets, private customer records, or unrelated proprietary files into prompts just to "fill context." Treat long-session conclusions as provisional when the repository, dependencies, docs, or issue state may have changed. Keep implementation authority separate from analysis: ask for review before applying broad file edits, running risky commands, or changing public project state. |
| Privacy notes | ✓A channel is an MCP server you configure with your own credentials; tokens are stored locally (for example ~/.claude/channels/telegram/.env). Incoming and outgoing messages pass through the third-party platform (Telegram, Discord, Apple Messages); review what data flows there. The iMessage channel reads your Messages database and requires Full Disk Access; understand that exposure before enabling it. | ✓Agent view row summaries are generated by a Haiku-class model from session activity; treat summaries as operational hints, not sanitized exports for public sharing. Sessions started under different project directories appear in the default global list unless you filter with claude agents --cwd. Peek panels and attach mode expose full transcripts that may contain proprietary code, tokens, or customer data. | ✓Memory files persist across sessions and may be committed if stored in project scope. Compaction summaries can retain file paths, issue titles, and internal URLs—review before sharing session exports. Clear memory when offboarding contractors or rotating shared machines. | ✓Long coding sessions can accumulate source code, file paths, stack traces, logs, issue details, usernames, and internal decisions. Durable memory and project instructions can persist sensitive details longer than an ordinary prompt, so store only stable, non-secret facts there. Handoff summaries should mention decisions and next steps without copying full private logs, tokens, customer reports, or unnecessary code excerpts. |
| Prerequisites |
|
|
|
|
| Install | — | — | — | — |
| Config | — | — | — | — |
| Citations | ||||
| Claim | Unclaimed | Unclaimed | Unclaimed | Unclaimed |
Source-backed guides for putting this to work.
Set autoMode.hard_deny rules to block risky actions in auto mode.
Loading live community signals…
A short, calm digest of reviewed Claude resources. Unsubscribe any time.