The Agent Skills Directory

[PROMPT_INJECTION]: The skill contains instructions and templates for processing untrusted data from Telegram users, creating an indirect prompt injection surface.
Ingestion points: The telegram_message trigger and associated handlers described in SKILL.md ingest text, photos, documents, and voice messages from external Telegram users into the agent context.
Boundary markers: The provided n8n workflow and AI integration templates do not include boundary markers (e.g., delimiters) or instructions to ignore embedded commands when interpolating {{ $json.message.text }} into LLM prompts.
Capability inventory: The skill defines capabilities for bot development and message handling, utilizing tools like telegram_send_message and telegram_send_photo to act upon processed input.
Sanitization: The example templates lack evidence of sanitization, escaping, or filtering for external content before it is processed by AI models.

telegram-bot