MCP / AI
Tool Catalog

MCP Tool Catalog

These are the tools the MCP server exposes to AI clients. Access reflects workspace RBAC: WS member means any role in the workspace, WS editor+ means editor or owner, and Admin only means a platform admin.

ToolParametersAccessDescription
list_workspacesAll usersList workspaces the user belongs to
list_ideasworkspace_id, status?, priority?WS memberList ideas in a workspace
create_ideaworkspace_id, title, description?, tags?WS memberCreate a new idea
update_ideaidea_id, status?, priority?, tags?WS memberUpdate idea fields
list_postsworkspace_id, status?WS memberList posts with pipeline status
get_postpost_idWS memberGet post content and metadata
generate_postidea_id, prompt_id?WS editor+Generate a post draft from an idea
iterate_postpost_id, instructionWS editor+AI iteration on post content
get_pipeline_statusworkspace_idWS memberPost counts by status
get_scheduleworkspace_id, weeks?WS memberUpcoming scheduled posts
list_actionsworkspace_id, status?WS memberList action requests
create_actionworkspace_id, title, assignee_id, post_id?WS editor+Create an action request
list_promptsworkspace_idWS memberList prompt templates
list_examplesworkspace_idWS memberList example posts
admin_list_usersstatus?Admin onlyList platform users
admin_approve_useruser_idAdmin onlyApprove a pending user

Implementation

Tools are grouped by domain under mcp/src/lkwiz_mcp/tools/ (workspaces.py, ideas.py, posts.py, genai.py, schedule.py, actions.py, prompts.py, examples.py, admin.py). Each tool delegates to the same business logic as its REST counterpart, so behavior stays in sync across the web app and MCP. See the project's MCP ↔ API parity check (scripts/check_mcp_parity.py).