The Agent Skills Directory

[PROMPT_INJECTION] (SAFE): The instructions in SKILL.md are designed to ground the AI's responses in the provided reference materials. There are no attempts to bypass safety filters or override system constraints.
[DATA_EXPOSURE] (SAFE): No sensitive file access or network exfiltration patterns were found. The skill discusses security concepts such as 'bot token leakage' and 'privacy settings' as educational topics for users, rather than performing any data access itself.
[REMOTE_CODE_EXECUTION] (SAFE): No remote code execution or external downloads are present. The references to third-party bots (e.g., Combot, Rose, Shieldy) are mentions of existing Telegram services for user implementation, not scripts to be executed by the agent.
[INDIRECT_PROMPT_INJECTION] (LOW): The skill ingests untrusted user data in the form of Telegram configuration queries. However, its capabilities are restricted to generating text-based advice and diagnoses. It lacks the ability to execute commands, write files, or perform network requests, which significantly mitigates the risk of indirect injection attacks.
[OBFUSCATION] (SAFE): All content is provided in clear-text Markdown. No hidden characters, Base64 encoding, or homoglyphs were detected.

telegram-mastery