MUSTer MCP Server

Instructions

Implementation Reference

• main.py:31-44 (registration)

  Registration of the MCP tool 'get_course_content' defining its name, description, and input schema.
  Tool( name="get_course_content", description="Get all assignments/resources for a course by exact name (call get_all_courses first).", inputSchema={ "type": "object", "properties": { "course_name": { "type": "string", "description": "Course name exactly as returned by get_all_courses", } }, "required": ["course_name"], }, ),
• main.py:35-43 (schema)

  Input schema definition for the tool, specifying 'course_name' as required string parameter.
  "type": "object", "properties": { "course_name": { "type": "string", "description": "Course name exactly as returned by get_all_courses", } }, "required": ["course_name"], },
• main.py:116-141 (handler)

  Primary MCP tool handler: resolves course by name, fetches content via MUSTerClient, formats as list of dicts with error handling.
  def tool_get_course_content(course_name: str) -> List[Dict[str, Any]]: try: all_courses = muster_client.get_courses() match = next((c for c in all_courses if c.name == course_name), None) if not match: return [{"error": f"No courses found matching '{course_name}'. Please call get_all_courses first."}] try: assignments = muster_client.get_course_content(match.url) return [ { "name": assignment.name, "type": assignment.type, "url": assignment.url, "course_name": match.name, "course_url": match.url, } for assignment in assignments ] except Exception as e: return [{"error": f"Failed to get content from course '{match.name}': {str(e)}"}] except Exception as e: return [{"error": f"Failed to get course content: {str(e)}"}]
• MUSTerClient.py:315-374 (helper)

  Core helper method implementing the scraping logic: loads course page, iterates sections and activities, determines types, builds Assignment list.
  def get_course_content(self, course_url: str) -> List[Assignment]: """Get all assignments and content from a specific course.""" self._ensure_driver() self.heartBeat() if not self.logged_in: if not self.login(): raise Exception("Login required to get courses.") try: self.driver.get(course_url) WebDriverWait(self.driver, 10).until( EC.presence_of_element_located((By.CSS_SELECTOR, ".course-content, .topics, li.section")) ) assignments = [] sections = self.driver.find_elements(By.CSS_SELECTOR, "li.section.main") for section in sections: try: section_title_element = section.find_element(By.CSS_SELECTOR, ".sectionname span a, .sectionname a") section_title = section_title_element.text.strip() if section_title.lower() in ["general", ""]: continue activities = section.find_elements(By.CSS_SELECTOR, ".activity") for activity in activities: try: activity_link = activity.find_element(By.CSS_SELECTOR, ".activityinstance a") instance_name_element = activity_link.find_element(By.CSS_SELECTOR, ".instancename") activity_name = " ".join(instance_name_element.text.strip().split()) activity_url = activity_link.get_attribute('href') activity_type = "resource" if "forum" in activity.get_attribute("class"): activity_type = "forum" elif "assign" in activity.get_attribute("class"): activity_type = "assignment" elif "quiz" in activity.get_attribute("class"): activity_type = "quiz" elif "resource" in activity.get_attribute("class"): activity_type = "file" if activity_name and activity_url: hierarchical_name = f"{section_title} > {activity_name}" assignments.append(Assignment( name=hierarchical_name, type=activity_type, url=activity_url, course=section_title )) except Exception: continue except Exception: continue self.heartBeat() return assignments except TimeoutException: print(f"Timeout while waiting for course content to load for url: {course_url}") return [] except Exception as e: print(f"An error occurred while getting course content: {e}") return []
• main.py:209-210 (handler)

  Dispatch logic in MCP server.call_tool() that routes calls to the tool_get_course_content handler.
  if name == "get_course_content": return _wrap_json(tool_get_course_content(args["course_name"]))