Atlas / MCP servers / bymercy / ramus-mcp

ramus-mcpSAFE

Name: ramus-mcp
Author: bymercy

mcp/bymercy/ramus-mcp

MCP server by bymercy · business process modeling · claude · claude desktop · Audited 2026-10-11

Claude's eyes and hands for Ramus IDEF0 models: read, see, check, edit and create .rsf files over MCP. Claude Desktop extension.

Verdict

SAFE

Grade

Trust score

89 /100

Exposed tools

28 13r · 13w · 2d

Transport

—

License

MIT

Stars

Overview

From the repository's own README, as read at the audited commit. Badges and raw HTML are left out.

Claude's eyes and hands for Ramus IDEF0 models

[](https://github.com/byMERCY/ramus-mcp/releases/latest) [](https://github.com/byMERCY/ramus-mcp/actions/workflows/tests.yml) [](LICENSE) [](https://modelcontextprotocol.io) [](#install-in-claude-desktop) [](pyproject.toml)

English · Русский

ramus-mcp lets Claude — or any MCP client — work with models made in Ramus, the IDEF0 business-process modelling tool. It opens a .rsf file, reads the activity tree and every arrow, and sees each diagram as a picture. It checks the model against the rules of IDEF0, changes it with boxes and arrows placed and routed the IDEF0 way, or builds a new model from nothing — and saves a .rsf that Ramus opens.

A model built from nothing — every frame is one tool call, drawn by the server itself.

What it does

Read from source at commit 3b05a437d7ccOBSERVED · 2026-10-11

Connect

Built from this server's own package name, version and transport as found in its source — not copied from anyone's documentation, so it cannot drift against a page we do not control.

claude-code

claude mcp add ramus-mcp -- uvx ramus-mcp

claude-desktop

{
  "mcpServers": {
    "ramus-mcp": {
      "command": "uvx",
      "args": [
        "ramus-mcp"
      ]
    }
  }
}

Exposed tools (28)

13 read · 13 write · 2 destructive. Blast radius: 2 tools can delete or overwrite — an agent that can be talked into calling a tool can be talked into calling this one.

Tool	Risk	Description
`add_activity`	write	Add an activity box to the decomposition of ``parent`` (its number,
`add_arrow`	write	Draw an arrow on a sheet - ``sheet`` is the number of the activity it decomposes (
`add_text`	write	Write a free text on a sheet - ``sheet`` as for add_arrow (
`check_layout`	read	How well a sheet is drawn - ``sheet`` is the number of the activity it decomposes
`check_model`	read	Check the model against the rules of IDEF0 - the whole model, or one sheet (``sheet``:
`create_model`	write	Create a new Ramus model file and open it - the way to start a model from nothing.
`current_model`	read	Say which model is open, if any.
`delete_activity`	destructive	Delete an activity box - ``activity`` is its number (
`delete_arrow`	destructive	Delete one arrow segment - its id is in get_diagram
`export_diagram`	write	Write one diagram - or every diagram of the model - to picture files, for a report or
`get_diagram`	read	Read one diagram as data: its boxes, the flows between them, and the free text on it.
`get_function_tree`	read	The whole activity tree: every box in the model, nested by decomposition.
`join_levels`	read	Join an arrow on a box to the same arrow on the frame of that box
`layout_sheet`	read	Lay a whole sheet out again - ``sheet`` is the number of the activity it decomposes
`list_diagrams`	read	List the model
`list_models`	write	Find Ramus model files (.rsf): in ``folder``, or else in the models folder set up for
`move_activity`	write	Move an activity box on its sheet, resize it, or both - ``activity`` is its number
`open_model`	write	Open a Ramus .rsf model file and make it the one every other tool reads and edits.
`rename_activity`	write	Give an activity box a new name. ``activity`` is its number (
`rename_flow`	write	Give a flow - what an arrow carries, shared by all its segments and levels - a new name.
`render_diagram`	read	Draw one diagram and return it as a PNG image, laid out as the model stores it.
`render_diagram_svg_text`	read	The same picture as render_diagram, as an SVG document (text) - for a person to save or
`save_model`	write	Write the model with every change made so far.
`set_purpose`	write	Write the model
`style_activity`	read	Give a box a look - ``activity`` is its number (
`style_arrow`	read	Give an arrow a look - ``segment`` is any of its segment ids (get_diagram with
`tidy_labels`	write	Move only the arrow names on a sheet that are in the way - on a box, a line or another
`tidy_sheet`	read	Lay out a sheet

Trust audit

SAFEgrade B · trust 89/100 Nothing in the source contradicts what it says it does. Grade A is reserved for packages that have also passed the behavioural sandbox.

Layer	What it checks	Result
L0	Provenance & inventory	PASS
L1	Static analysis of the code	PASS
L2	Instruction surface (what it tells the agent)	PASS
L3	Class-specific surface	WARN
L4	Behavioural (sandbox)	SKIPPED

What the source does

Filesystem: declared (2 observation(s))
Network: none-observed
Shell: none-observed
Dependencies: not all pinned
Secrets in source: none-found

Findings (12)

MEDIUMFilesystem / path · mcp.destructive_tools · CWE-22, CWE-59

delete_activity, delete_arrow

Why it matters. 2 tool(s) can delete or overwrite

Fix. prefer a read-only mode or scoped tokens; the page states the blast radius

LOWInventory / provenance · inv.hidden_file · CWE-1104

.mcpbignore

.mcpbignore

Why it matters. hidden member outside the usual dotfiles

Fix. review its purpose

LOWObfuscation / stealth · obf.decode_call · CWE-506, CWE-94

src/visual_data.py:46

return bytes.fromhex(text.strip()).translate(_UNMASK)

LOWObfuscation / stealth · obf.decode_call · CWE-506, CWE-94

tests/test_server.py:435

self.assertTrue(base64.b64decode(image.data).startswith(b"\x89PNG"))

LOWObfuscation / stealth · obf.decode_call · CWE-506, CWE-94

tests/test_visual_data.py:35

raw = bytes.fromhex(

LOWObfuscation / stealth · obf.decode_call · CWE-506, CWE-94

tools/ramus_validate.py:115

fh.write(base64.b64decode(c["data"]))