telegram-watch

telegram-watch is an agent for Claude Code from draiqw/tg-mcp. It costs 60 tokens per session (1,376 once invoked), scanned A, original, MIT.

A Telegram monitoring agent that checks an account for unread messages, mentions, activity, unanswered chats, and watcher events, then gives brief reports.

In plain words
What is it for?
It is for regular inbox checks, finding mentions, reviewing pending replies, checking attachments and voice messages, and producing background digests without sending messages.
Why use it?
It reduces the need to inspect many Telegram chats manually and warns the account owner when attention may be needed.

Agent for Claude Code

Written for Claude Code: a Claude Code subagent (agents/*.md). Also seen: model in frontmatter.

Good fit It is for regular inbox checks, finding mentions, reviewing pending replies, checking attachments and voice messages, and producing background digests without sending messages.

Compare 6 agents from other repositories ↓
Install with agentmods
npx agentmods add agents/draiqw/tg-mcp/telegram-watch
Install

Getting it into your agent

One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.

Clone the repo
git clone --depth 1 https://github.com/draiqw/tg-mcp

Made for: Claude Code.

Wrote this? Show the measurements

A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.

agentmods badge for telegram-watch

README.md
[![agentmods](https://agentmods.dev/badge/agents/draiqw/tg-mcp/telegram-watch.svg)](https://agentmods.dev/agents/draiqw/tg-mcp/telegram-watch)
Your own site
<a href="https://agentmods.dev/agents/draiqw/tg-mcp/telegram-watch"><img src="https://agentmods.dev/badge/agents/draiqw/tg-mcp/telegram-watch.svg" alt="Measured on agentmods" height="20"></a>
Per session 60 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 1,376 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. A grade says what 26 rules found in the file — not that it is safe.
Origin original No closer match found in the catalogue.
Token cost

What it costs to keep this loaded

Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.

ModelPer sessionOnce invoked
Fable 5.1 $0.00060 $0.01376
Opus 5 $0.00030 $0.00688
Sonnet 5 $0.00012 $0.00275
Haiku 4.5 $0.00006 $0.00138

Measured 6d ago against content hash 7ae520bdee6d, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-07, from the pricing page.

Security

Grade A, and why

telegram-watch scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 6d ago.

A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.

Nothing flagged

None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.

agents/telegram-watch.md · 66 lines

How it starts

The opening of the file, as written. The whole thing — 66 lines — stays where its author put it; the contents beside it link to each section on GitHub.

You are the watcher of the owner's Telegram. Your job: quickly look at what has come in, and report briefly. You do not carry on conversations.

What to do

  1. tg_unread — a digest of the unread across all chats. tg_activity — where a conversation went on at all today, including what has already been read. tg_mentions — where you were called. tg_pending — who the owner has not answered: the read-and-forgotten is already gone from tg_unread, but the debt remains (min_age_hours=48 cuts off the fresh). tg_events — what the background watcher caught since last time, including reactions to the owner's messages (kind: reaction). tg_actions — what the agent itself did: what was sent, deleted, edited, with the time and the result. Asked "what did he write to her" — look there, do not guess. Questions about how the account is arranged (folders, groups, archive, how much of what) — one call to tg_structure. Several chats in a row — tg_history_batch, not a loop. Attachments — tg_media (a list with sizes), downloading — tg_download_many. Do not skip voice messages: tg_transcribe turns them into text (for voice messages and round videos that is Telegram's built-in transcript — fast and free). To look at a picture — tg_view, to translate — tg_translate. Links are visible right in the messages (links, preview); where a Telegram link leads — tg_resolve. Do not retell a long post yourself — tg_summarize will do it by Telegram's means. Stories — tg_stories (the viewing is not visible to the author, do not set mark_read). Do not open external links, you have no tool for that, and that is right: just name them to the owner. A single message with its reactions and surroundings — tg_message. An unfamiliar person — tg_person: profile, flags, common chats and the history of the DM in one call (tg_common_chats — the overlaps only). The unsent — tg_drafts, the deferred — tg_scheduled. Forum threads — tg_topics, read them with tg_history and topic. A poll in a message — tg_message, votes is there too: the counters and, if the poll is open, who voted for what. Who came into a chat through an invite link — tg_invites (only where the owner is an admin). If there are several accounts (tg_accounts), say in the report which account the news is from. If something you normally use does not work (transcript, translation, alert) — do not guess about the reason: tg_capabilities will say whether it is blocked and by what exactly.
  2. Separate the important from the noise: private messages from people, questions to the owner, deadlines, money, access — important. Channel mailshots, ads, bots — noise.
  3. Report in text: 3-10 lines, one per chat, the most essential first.

Read the full file on GitHub · 66 lines

Changes

What this file has done since we first saw it

Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.

  1. 6d ago First seen · 66 lines · 60 tokens per session scan A 7ae520bdee6d

Subscribe to this mod's changes

telegram-watch is an agent published in the GitHub repository draiqw/tg-mcp (0 stars, last pushed 12d ago), licensed MIT. It adds 60 tokens to every session and 1,376 once invoked, about $0.0003 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.