← Registry

Design

flowgo-map.com

An MCP server for creating, editing, and sharing Flowgo diagram maps in an in-memory workspace.

1 endpoint22 known toolsFirst detected October 19, 2026Last detected October 19, 2026

ENDPOINT 1

https://flowgo-map.com/api/mcp

No auth detected

MCP server metadata

Name
flowgo
Version
cloud
Capabilities
resourcestools
Server instructions

flowgo is a mind-map / flowchart / whiteboard editor backed by a plain-text .flowgo file. Reach for these tools whenever the user asks to map this out, whiteboard, sketch, diagram, lay it out, or draw a mind map, system map, or canvas — flowgo renders live and is interactive, unlike a static mermaid code block. Each file is a tree of maps; each map holds positioned boxes (labelled nodes), edges (undirected links between boxes), free-floating text labels, static lines, and freehand strokes. WHY FLOWGO OVER A STATIC DIAGRAM CODE BLOCK Every node can contain its own submap — double-click in, and that node's insides are a full nested canvas at "/<node_id>". This gives zoomable, drill-downable maps (system → component → function) that a flat flowchart-as-text format can't express, and the result is a live editable canvas the user can share and keep working in, not a rendered-once image. MAPS AND SUBMAPS A map is addressed by path. "/" is the root. "/<node_id>" is the inside of a node — a submap. "/<node_id>/<inner_node_id>" is two levels deep. Submaps are created implicitly the first time you write to a new path; you don't need a "create_map" call. To navigate, pass the path to any tool. Nodes carry the same ids across maps because each map's id space is independent. COORDINATES x, y are in CSS-pixel data space. Origin is top-left; +x is right, +y is down. Nodes render roughly 120-180px wide and 36-44px tall depending on label length and font, so space them by at least 200px horizontally and 80px vertically to avoid overlap. A reasonable map fits inside a ~1600x1200 canvas; large maps work but the GUI will need pan/zoom. Always pass distinct coordinates — multiple items at (0, 0) will pile up. STYLING (1-9 SCALES) - palette: 1=default (white node, black text), 2=blue, 3=purple, 4=green, 5=yellow, 6=red, 7=orange, 8=gray, 9=black/inverted (black bg, white text). Applies to nodes, edges, texts, lines, strokes. - font (nodes, texts): 1=default 14px, scales up to 9 ≈ 56px. - style (lines only): 1=straight, 2=smooth bezier, 3=orthogonal right-angle elbows. - shape (nodes only): 0=rectangle (default), 1=hexagon, 2=circle, 3=triangle. Non-rectangles render at fixed sizes in the GUI (hexagon 240x208, circle 208x208, triangle 240x208) and are not resizable; hexagons additionally snap onto a hex lattice near other hexagons and never overlap. EDGE HANDLES fromHandle/toHandle pin the connection to a specific side or corner of the source/target node: t (top), r (right), b (bottom), l (left), tl, tr, bl, br. Omit both to let the renderer auto-pick the nearest pair. Edges are undirected — add_edge from A to B is the same edge as B to A; update_edge / delete_edge match in either order. EDGE LABELS An edge carries an optional label drawn at its midpoint — "depends on", "triggers", "owns". Set it with add_edge/update_edge 'label'; pass an empty string to update_edge to remove it. Label the relationship whenever the connection's meaning isn't obvious from the two node names: a map that shows only THAT things connect, never HOW, throws away most of what a system map is for. WHEN TO USE WHICH ENTITY - node: the labelled unit of the conceptual graph (the add_box/update_box/delete_box tools keep their historical names — nodes were formerly called boxes). Use for things that have meaning and connect. - edge: a connection between two nodes. Always between nodes — not to text or lines. Carries an optional label naming the relationship. - text: a free-floating annotation. Use for callouts, headers, labels that aren't graph nodes. - line: a static segment with optional control points. Use for arrows, dividers, geometric shapes. - stroke: freehand polyline. Use for sketchy annotations; agents rarely need this. - anchor (a flag on one node per map): marks the recenter target the GUI scrolls to on load. Optional. ACCOUNTS If an "authenticate" tool is listed, this server can link the session to the user's flowgo account — call it when they ask to save a map to their account or sign in, and hand them the URL it returns. Without it, shared maps are anonymous but still reachable by link. GRANULAR VS SET_STATE Prefer add_*/update_*/delete_* for edits. set_state rewrites the entire graph and runs strict validation; it's the right tool for bulk imports or wholesale layout changes, not for tweaking one node's color. For the full file-format reference (the .flowgo on-disk syntax), read the resource flowgo://about.

Known tools 22

start_workspace

Create a new in-memory workspace for this session and return its workspace_id.

Potential side effects
share

Persist the current workspace as an immutable snapshot via the configured webhook.

Inferred read-only
create_map

One-shot: turn complete .flowgo text into a public share URL.

Inferred read-only
authenticate

Link this session to the human's flowgo account, so maps you share or create are saved to it instead of being anonymous.

Potential side effects
get_state

Read and return the entire flowgo graph (every map with its nodes, edges, texts, lines, and strokes).

Inferred read-only
set_state

Replace the entire graph with the supplied object.

Inferred read-only
add_box

Add a node to the map at the given path.

Inferred read-only
update_box

Update a node's label, position, color, font size, shape, anchor flag, or explicit size.

Potential side effects
set_default_shape

Set the document-level default shape (the `defaultshape <n>` directive): the shape a canvas double-click creates in this file.

Inferred read-only
delete_box

Delete a node (and all incident edges plus its submap subtree).

Potential side effects
add_edge

Add an undirected edge between two nodes in the same map.

Inferred read-only
update_edge

Update the edge between two nodes: re-aim a handle, set the palette, or set/clear the midpoint label.

Potential side effects
delete_edge

Delete the edge between two node ids in the same map.

Potential side effects
add_text

Add a free-floating text annotation (header, callout, label).

Inferred read-only
update_text

Update a text label's content, position, color, or font size.

Potential side effects
delete_text

Delete a free-floating text label.

Potential side effects
add_line

Add a static line segment (arrow, divider, geometric shape).

Inferred read-only
update_line

Update a line's endpoints, color, render style, or control points.

Potential side effects
delete_line

Delete a static line segment.

Potential side effects
add_stroke

Add a freehand brush stroke (sketchy annotation).

Inferred read-only
update_stroke

Update a stroke's color or replace its full point list (useful for translating the whole stroke).

Potential side effects
delete_stroke

Delete a freehand brush stroke.

Potential side effects

CONNECT WITH APPROVAL

Client installation

Review this server and its permissions before adding it. Secret placeholders must be set locally.

Codex

~/.codex/config.toml

[mcp_servers.flowgo]
url = "https://flowgo-map.com/api/mcp"
enabled = true
Claude Code

.mcp.json

{
  "mcpServers": {
    "flowgo": {
      "type": "http",
      "url": "https://flowgo-map.com/api/mcp"
    }
  }
}
Claude Desktop

Settings → Connectors → Add custom connector

Name: flowgo
Remote MCP URL: https://flowgo-map.com/api/mcp

Add this remote URL as a custom connector in Claude Desktop. Availability depends on the user plan and workspace policy.

Cursor

.cursor/mcp.json

{
  "mcpServers": {
    "flowgo": {
      "url": "https://flowgo-map.com/api/mcp"
    }
  }
}
Visual Studio Code

.vscode/mcp.json

Add to Visual Studio Code
{
  "servers": {
    "flowgo": {
      "type": "http",
      "url": "https://flowgo-map.com/api/mcp"
    }
  }
}
Generic MCP

Client-specific MCP configuration

{
  "name": "flowgo",
  "transport": "streamable-http",
  "url": "https://flowgo-map.com/api/mcp"
}
MCP Inspector

Run the official MCP Inspector locally and enter the indexed Streamable HTTP endpoint.

TRUST AND VERIFICATION EVIDENCE

Loading Trust v2 evidence…

Checking the associated registrable domain. The BuiltWith key remains server-side.

Indexed

Evidence is source-attributed and does not guarantee that a third-party server is safe. Risk labels are conservative metadata heuristics.