custom-telegram-bot

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

• Third-party content exposure detected (high risk: 0.75). The required runtime workflow for get_updates ingests inbound Telegram update text (and optionally downloads media) from other chat participants, so outsider-authored message body text can be returned to the agent as readable JSON/prose and thus enter the LLM context (indirect prompt injection risk).

MEDIUM W012: Unverifiable external dependency detected (runtime URL that controls agent).

• Potentially malicious external URL detected (high risk: 0.90). This skill explicitly instructs runtime calls to the AgentPMT MCP endpoint to fetch live schema/instructions (e.g., via https://api.agentpmt.com/mcp/ using the agentpmt-tool-search-and-execution get_instructions/get_schema actions), and those fetched instructions are treated as authoritative and directly control the tool's prompts/behavior.