Posts MCP Tools

Three read-only tools give an AI client (Claude Desktop, Claude Code, or any MCP-speaking assistant) visibility into a workspace's post pipeline — list posts, fetch one, or get pipeline counts. There is currently no MCP tool to create, edit, or transition a post; use the REST API or the app UI for writes.

For the underlying business rules, see Features — Posts. This page is based on specs/mcp/posts.md in the repo — see that spec for the exact JSON shapes and DynamoDB access pattern behind each tool.

Access model

Every tool call is scoped to a workspace_id and enforces membership the same way the REST API does — a direct DynamoDB lookup on WS#{workspace_id} / MEMBER#{user_id}. If the authenticated MCP user isn't a member, the tool raises and the client sees a tool error, not an empty result.

⚠️

All three tools return a JSON string, not a structured object — parse it before use. This is consistent across all lkwiz_mcp post tools.

Tools

Tool	Parameters	Description
`list_posts`	`workspace_id`, `status?`	List posts in a workspace, optional status filter
`get_post`	`post_id`, `workspace_id`	Fetch one post's content and metadata
`get_pipeline_status`	`workspace_id`	Count posts per pipeline status

`list_posts`

list_posts(workspace_id: str, status: str = "") -> str

Returns every post under the workspace, optionally filtered to an exact status match ("draft", "in_review", "approved", "scheduled", "published", "archived", or "revision_requested" — the filter accepts any of these; see the note on get_pipeline_status below for why one of them doesn't show up in the counts tool).

{
  "posts": [
    {
      "post_id": "post_abc",
      "workspace_id": "ws_123",
      "idea_id": "",
      "title": "Why observability matters",
      "content": "<p>...</p>",
      "status": "draft",
      "scheduled_at": "",
      "created_at": "2026-07-11T10:00:00+00:00",
      "updated_at": "2026-07-11T10:00:00+00:00"
    }
  ]
}

⚠️

The idea_id field is always "" in practice — posts are stored with a source_idea_id attribute, not idea_id, so this projection never finds a value to read. Look up the post's linked idea via the REST API (GET /api/posts/{id}, field source_idea_id) instead.

`get_post`

get_post(post_id: str, workspace_id: str) -> str

Returns a single post (same shape as one item in list_posts's posts array, but as the top-level object — not wrapped). Raises if the post doesn't exist in that workspace.

`get_pipeline_status`

get_pipeline_status(workspace_id: str) -> str

The MCP equivalent of the kanban board's column headers — how many posts are in each stage.

{
  "workspace_id": "ws_123",
  "pipeline": {
    "draft": 3,
    "in_review": 1,
    "approved": 2,
    "scheduled": 1,
    "published": 5,
    "archived": 0
  }
}

⚠️

revision_requested posts are not counted here — the counter map only has the six keys shown above. A post in revision_requested is fetched from DynamoDB along with everything else, but silently skipped when tallying because its status string isn't one of the map's keys. If you need a revision_requested count, use list_posts(status="revision_requested") and count the results yourself.

GenAI MCP Tools Schedule MCP Tools