telegram-telethon

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 1.00). The non-interactive setup examples show passing api_id, api_hash, verification codes, and passwords directly on the command line (e.g. --api-hash, --code, --password), which requires the LLM to include secret values verbatim in generated commands/outputs and therefore creates an exfiltration risk.

CRITICAL E005: Suspicious download URL detected in skill instructions.

• Suspicious download URL detected (high risk: 0.80). Several short/obscure domains in the list (https://x.io, https://b.io, https://a.io, https://x.y) are URL-shortener / minimal domains or non-standard hosts that can obscure redirect destinations and are commonly used to hide malicious downloads; the remaining links (api.groq.com, my.telegram.org, example.com, github.com/glebis/claude-skills, t.me/klodkot) are legitimate API, vendor, placeholder, or repository references found in the skill files.

MEDIUM W011: Third-party content exposure detected (indirect prompt injection risk).

• Third-party content exposure detected (medium risk: 0.30). The Telegram daemon reads outsider-authored message free text from Telegram events (event.message.text) and, when a trigger matches, forwards the captured/full text to Claude via claude_bridge.send (handlers.py _match_message → MessageHandler.handle → ClaudeBridge.send → claude -p prompt).