telegram-bot-development
Pass
Audited by Gen Agent Trust Hub on Nov 7, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTIONOBFUSCATION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The script
scripts/tgdocs.pydownloads official Bot API documentation fromcore.telegram.organdtelegram.org. These are well-known, trusted domains for the service being managed, and the downloads are limited to documentation files (.md and .txt). - [COMMAND_EXECUTION]: The script
scripts/find_method_support.pyusessubprocess.runto execute system commands such asgo envand invokes the current Python interpreter (sys.executable). These operations are used locally to discover library installation paths for code analysis and do not involve shell execution of untrusted input. - [INDIRECT_PROMPT_INJECTION]: The skill provides extensive guidelines on identifying and mitigating indirect prompt injection. It instructs developers to treat external text (from users, channels, or other bots) as data rather than instructions and recommends the use of boundary markers and sanitization when integrating with LLMs.
- [OBFUSCATION]: The script
scripts/find_method_support.pycontains a Unicode Line Separator character (U+2028). This character is used non-maliciously and is explicitly documented in a code comment that explains the script's logic for handling various line-ending characters during text parsing.
Audit Metadata