The sprint summary and the task-detail tools are what save time — one call gets the sprint's status distribution instead of counting cards. Semantic search is the one capability with an extra dependency behind it: it queries a vector database you stand up separately, so treat it as opt-in.
A server over the Goodday project-management API, grouped into project, task, sprint, user and search tools.
- Projects: get_projects with archived and root-only filters, get_project, create_project and get_project_users
- Tasks: get_project_tasks, get_user_assigned_tasks, get_user_action_required_tasks, get_task, and get_task_details with subtasks, custom fields and full metadata
- Task activity: get_task_messages for the comment thread, add_task_comment to add one, update_task_status to move a task on, create_task to open one with assignments, dates and priority
- Sprints: get_goodday_sprint_tasks by project and sprint, and get_goodday_sprint_summary for status distribution and key metrics
- Users: get_users and get_user
- get_goodday_smart_query answers plain-language project questions; search_goodday_tasks does semantic search over tasks; search_project_documents and get_document_content cover project documents
A Goodday API token generated from your organisation's settings, handed to the server as an environment value. Python. Semantic task search expects a vector-search endpoint you host yourself; every other tool works without it.
One command plus a key — pip install goodday-mcp, then supply credentials
