aiogram-mcp

Instructions

Implementation Reference

• aiogram_mcp/tools/chats.py:156-191 (handler)

  Implementation of the unban_user MCP tool handler.

  async def unban_user(
      chat_id: int,
      user_id: int,
      only_if_banned: bool = True,
  ) -> UnbanUserResult:
      """Unban a previously banned user."""
      if not ctx.is_chat_allowed(chat_id):
          result = UnbanUserResult(ok=False, error=f"Chat {chat_id} is not allowed.")
          if ctx.audit_logger:
              ctx.audit_logger.log(
                  "unban_user",
                  {"chat_id": chat_id, "user_id": user_id},
                  result.ok,
                  result.error,
              )
          return result
  
      try:
          if ctx.rate_limiter:
              await ctx.rate_limiter.acquire()
          await ctx.bot.unban_chat_member(
              chat_id=chat_id,
              user_id=user_id,
              only_if_banned=only_if_banned,
          )
          result = UnbanUserResult(ok=True, user_id=user_id)
      except (TelegramBadRequest, TelegramForbiddenError) as exc:
          result = UnbanUserResult(ok=False, error=str(exc))
  
      if ctx.audit_logger:
          ctx.audit_logger.log(
              "unban_user",
              {"chat_id": chat_id, "user_id": user_id},
              result.ok,
              result.error,
          )

• aiogram_mcp/tools/chats.py:34-35 (schema)

  Result schema for unban_user tool.

  class UnbanUserResult(ToolResponse):
      user_id: int | None = None

• aiogram_mcp/tools/chats.py:153-153 (registration)

  Registration logic for unban_user tool within register_chat_tools.

  if allowed_tools is None or "unban_user" in allowed_tools: