Telegram Bot API

Build Telegram bots with correct API calls, message formatting, keyboards, and webhook setup.

This skill is an instruction-only Telegram Bot API guide and is internally consistent. Key points before installing/using: 1) The skill may ask to save your bot token to ~/telegram-bot-api/bots/{name}.md — only allow this if you trust the environment and the machine's filesystem; prefer using a secrets manager or environment variables if you prefer. 2) Never paste your bot token into public chat or logs; if you suspect it was exposed, regenerate it with @BotFather. 3) Webhook examples require exposing an HTTPS endpoint (ngrok/cloudflared); be careful when exposing localhost to the public. 4) The skill's examples reference ${TOKEN} but it does not require any platform-level credentials — the token is a Telegram credential you provide. 5) Because this is instruction-only, there is no installer, no external downloads, and the repo contained only documentation — that reduces risk. If the skill later asked to run downloaded code, contact external endpoints other than api.telegram.org, or required unrelated credentials (AWS, GitHub tokens, etc.), reconsider or mark as suspicious.

Type: OpenClaw Skill Name: telegram-bot-api Version: 1.0.0 The skill is designed to assist with Telegram Bot API interactions, which inherently involves handling sensitive data like bot tokens and making network requests. The skill explicitly states that all external communication is directed solely to `https://api.telegram.org` and no other endpoints, as detailed in `SKILL.md`. Furthermore, `setup.md` provides clear instructions for the AI agent to ask for user permission before saving bot tokens locally to `~/telegram-bot-api/bots/{botname}.md` and to avoid displaying them. The use of `curl` and other command-line tools (like `ffmpeg`, `openssl`, `ngrok`, `cloudflared`) is for legitimate, transparent purposes related to bot development and webhook setup. There is no evidence of malicious intent, unauthorized data exfiltration, or deceptive prompt injection.