Claude Code Telegram Kit

Not another Telegram bridge. Anthropic's official Claude Code Channel keeps inbound. This kit fixes the two things it does not do: Markdown that survives Telegram's parser, and resetting context from your phone.

Research-preview infrastructure. Review the security model before connecting it to a machine with valuable data.

Official Channel	With this kit

The same Markdown document, both paths. The official reply tool defaults to format: "text", so markup arrives literal; its markdownv2 mode shifts MarkdownV2 escaping onto the model, where one missed character fails the send. send_reply takes the document unescaped and picks the transport itself. (Figures are rendered from both paths, not device screenshots.)

Why this exists

Every other "Claude Code + Telegram" project replaces the official Channel: its own poller, its own session management, its own pairing. This one does not. Inbound polling, sender pairing, attachments, and permission relay stay with Anthropic's plugin. The kit adds two bounded outbound/control capabilities beside it, without a second getUpdates consumer:

• Telegram Renderer MCP — one canonical send_reply(raw Markdown) tool with deterministic Rich Message vs MarkdownV2 routing, permanent-only fallback, and 👀 → 👍/👎 processing reactions.
• Session Control MCP — an approval-gated /reset path that finalizes the confirmed acceptance reaction, then hands execution to a root-owned, fail-closed local reset helper that PID 1 executes.

Both gaps are open upstream. This kit is the interim answer:

• anthropics/claude-code#39684 — no way to clear or reset context remotely
• anthropics/claude-code#36622 and claude-plugins-official#774 — requesting a MarkdownV2 parse_mode

Quickstart

Requires the official telegram@claude-plugins-official plugin already paired and working.

git clone https://github.com/project-tharsis/claude-code-telegram-kit
cd claude-code-telegram-kit
bun install --frozen-lockfile
bun run check

sha=$(git rev-parse HEAD)
python3 scripts/deploy_local.py install --repo . --ref "$sha" --bun "$(command -v bun)"

Then copy examples/.mcp.json, examples/telegram-settings.json, and examples/CLAUDE.md into your Claude project, replacing USER with your own paths. Merge examples/access-ux.json into the official Channel's access.json to enable the initial 👀 acknowledgement. Send a message with a GFM table; the renderer should report mode: rich and replace 👀 with 👍.

The renderer works on its own. /reset additionally needs the root helper, installed separately by the exact-commit procedure in the session-control README.

For production deployment, rollback, and verification, follow the operations runbook rather than this section.

Architecture

Telegram
  -> telegram@claude-plugins-official     # sole inbound poller
  -> Claude Code
     -> telegram-renderer MCP              # bounded outbound rendering
     -> session-control MCP                # bounded reset scheduling
        -> systemd transient unit
        -> root-owned session reset helper

The renderer and control MCPs reuse the official Channel's token and access.json authority. They require dmPolicy: allowlist, secure 0600 state files, and exact destination membership.

Design invariants

These five define the blast radius:

• One Telegram getUpdates consumer per bot token.
• No arbitrary Bot API method tool.
• No arbitrary shell command tool.
• Timeouts, 429s, 5xx responses, and unknown outcomes never trigger a resend.
• PID 1 owns reset execution before the Claude process is terminated.

The complete set is in docs/design-invariants.md.

Repository layout

packages/
  shared/                  Telegram authority validation
  telegram-renderer-mcp/   Markdown renderer and MCP server
  session-control-mcp/     Reset controller, MCP server, root helper
examples/                  Generic Claude, MCP, systemd, and reset config
scripts/                   Versioned local install and rollback

Requirements

• Linux with systemd and procfs mounted at /proc
• Claude Code 2.1.234 or newer
• Bun 1.3.14 or newer
• Python 3.11 or newer
• Anthropic's official telegram@claude-plugins-official plugin

Installation model

Do not run production from a mutable development checkout. Install an exact commit into a versioned release directory:

~/.local/share/claude-code-telegram-kit/
  releases/<git-sha>/
  current -> releases/<git-sha>
  previous -> releases/<previous-sha>

scripts/deploy_local.py extracts a Git archive with a Python 3.11-compatible no-link/no-traversal extractor, installs production dependencies, verifies the release receipt, and atomically swaps current/previous. It never installs root-owned files.

python3 scripts/deploy_local.py status
python3 scripts/deploy_local.py rollback

Keep Telegram credentials and allowlists under Claude's state directory, and keep reset configuration root-owned under /etc/claude-code-telegram-kit/.

Session reset

The local recovery authority is:

sudo claude-code-session-reset --config /etc/claude-code-telegram-kit/reset.json

The optional Telegram /reset command is a thin MCP front end. It cannot recover a Claude process that is already unable to receive messages; keep the local helper available as the break-glass path.

Development

bun install --frozen-lockfile
bun run check
bun audit

Security

Read SECURITY.md before deployment. Never commit bot tokens, chat IDs, transcripts, service-specific paths, or live reset configuration.

Project status

The code is extracted from a live, verified deployment, then generalized into a clean-room public repository. APIs may change before 1.0.0.

The initial release is source-only. Workspace packages are marked private and are not published to npm; install from an exact Git commit with the versioned deploy script.

License

Apache-2.0. See LICENSE, NOTICE, and THIRD_PARTY_NOTICES.md. Release procedure: RELEASING.md.

This project is independent and is not endorsed by Anthropic or Telegram.