telegram
Pass
Audited by Gen Agent Trust Hub on Sep 22, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill provides patterns for processing untrusted data from Telegram updates, including messages, callback queries, and inline queries, as documented in
references/bot-api.mdandreferences/aiogram-patterns.md. While the skill includes instructions for authentication verification, the handling of user-generated content presents a vulnerability surface where malicious inputs could influence agent behavior.\n - Ingestion points: update payloads (e.g.,
Message.text,CallbackQuery.data,InlineQuery.query) processed by the bot handlers.\n - Boundary markers: The skill advocates for hash verification for authentication data but does not specify delimiters or boundary markers for general message text processed by an AI agent.\n
- Capability inventory: Handlers have the capability to send/edit messages, interact with databases (PostgreSQL, Redis), and perform network operations via the Telegram Bot API.\n
- Sanitization: The skill provides correct implementation examples for HMAC-SHA-256 signature verification for web app and login data.
Audit Metadata