AgentSkillsCN

reader-telegram

通过 URL 获取 Telegram 频道帖子内容。当用户提供了 Telegram 的 URL（t.me、telegram.me），或请求阅读 Telegram 频道中的某条帖子时，可使用此功能。此外，当出现以下情况时也应触发该功能：(1) 用户分享 t.me 链接；(2) 用户说“读一读这条帖子”“Telegram 上这条帖子讲了什么”“去 Telegram 上看看”；(3) 需要对 Telegram 帖子进行分析或摘要总结。返回的帖子文本会保留内嵌链接，并附带作者、日期、浏览量及互动数据。

SKILL.md
--- frontmatter
name: reader-telegram
description: "Fetch Telegram channel post content by URL. Use when the user provides a Telegram URL (t.me, telegram.me) or asks to read a post from a Telegram channel. Also trigger when: (1) user shares a t.me link, (2) user says 'прочитай пост', 'что в этом посте в телеге', 'посмотри в телеграме', (3) a Telegram post needs to be analyzed or summarized. Returns post text with inline links preserved, author, date, views, and reactions."

Telegram Post Reader

Fetch public Telegram channel posts with full text and inline links preserved. Uses embed page parsing — no API keys needed.

Usage

bash
# Single post
uv run scripts/fetch_post.py "POST_URL"

# Multiple posts
uv run scripts/fetch_post.py "URL1" "URL2" "URL3"

The Iron Law

NEVER modify the post text. Output it exactly as received from the page. Summarize or analyze only AFTER presenting the original text to the user.

URL Formats Supported

  • •https://t.me/channel/post_id
  • •https://telegram.me/channel/post_id

Output Example

code
# Anton Vdovitchenko — Ai Agents
**URL:** https://t.me/aiwizards/397
**Date:** 2026-02-17T16:22:30+00:00
**Views:** 810
**Reactions:** ❤ 9 | 💯 8 | 👍 4 | 🔥 1

---

**Software On Demand.** Привет, друзья! ...
Помните историю про [ClawdBot](https://github.com/openclaw/openclaw)...

After Getting Post Content

  1. •Present the formatted output to the user
  2. •Then — and only then — analyze, summarize, or process if the user asked for it

Limitations

  • •Only public channels (private channels/groups will return errors)
  • •Media returned as URLs, not downloaded
  • •Depends on Telegram embed page availability
  • •No search capability — only fetches by URL

Related Skills

  • •reader-twitter — tweets and threads from X/Twitter
  • •reader-youtube — video transcript extraction from YouTube
  • •reader-jina — general fallback for JS-heavy web pages