MCP.Market

Tensorcad MCP server

by Filip-Pajalic·io.github.Filip-Pajalic/tensorcad·v0.1.21
No reviews yet

Design transformer LLM architectures and report their parameters, FLOPs, memory and cost

A88/100grade A
Write a reviewInstall
What users say
No reviews yet
Be the first
Safety scan
A88/100

full report

Adoption
Growing

0 stars149 downloads/wk

Reviews

Write one

Nobody has reviewed Tensorcad yet.

If you have run it, two minutes of your experience saves the next person an afternoon.

Tensorcad tools (19, 4 write)

write = sends, deletes, buys or posts

Read from the package source without running it. The installed server may list more.

  • tensorcad_analyze

    Parameters, FLOPs per token, KV cache, training and serving memory, decode throughput, training cost and

  • tensorcad_apply_ops

    Apply a batch of edits to a design. The batch is all-or-nothing: the first rejected operation aborts it

  • tensorcad_checkpoint

    Snapshot a design under a name you can come back to. Take one before an experiment so tensorcad_restore

  • tensorcad_diff

    What changed between two designs and what it cost: the symbols, blocks and wires that moved, then the

  • tensorcad_explain

    What one block is and what it contributes: its parameters as written and as evaluated, the shape on

  • tensorcad_generate_code

    Emit a runnable PyTorch module plus the design document. Without out_dir the file contents come back in

  • tensorcad_get_block

    One block of a design: its parameters as written and as resolved, the inferred shape on every port,

  • tensorcad_get_design

    Read a design. Use format "outline" (the default) first: it is the whole structure, symbol table and

  • tensorcad_import_hf

    Read a Hugging Face `config.json` into a design and save it in this session. Covers the Llama, Mistral,

  • tensorcad_list_designs

    List the designs this server has open, the built-in reference architectures you can start from,

  • tensorcad_mup

    The same design at several widths, with what to multiply the initialization and the learning rate by at

  • tensorcad_new_designwrite action

    Create a design from a reference architecture, or an empty one with just the B and T runtime symbols.

  • tensorcad_open_design

    Load a .tensorcad.json document from disk and return a design_id for it.

  • tensorcad_plan

    Every way of splitting the training across a cluster that fits, least demanding first. Prices data,

  • tensorcad_restorewrite action

    Put a design back. With a checkpoint_id it restores that snapshot; without one it undoes the most recent

  • tensorcad_save_designwrite action

    Write a design to disk as .tensorcad.json. Overwrites the file it was opened from unless a path is given.

  • tensorcad_scale

    Shrink a design towards a parameter budget while keeping its proportions, and save the result as a new

  • tensorcad_search_catalog

    Search the block catalog. Returns each block's parameter schema, port patterns and documentation,

  • tensorcad_validatewrite action

    Run every design rule: shape and symbol errors, kernel-friendly head dimensions, tensor-core multiples,

Public scan report

scanner v0.1.10 · 2026-11-04 · same rubric, same numbers if you re-run it

no findings
  • Code scan24 source files scanned25/25
  • –Live reliabilityno gateway calls yet and no remote to proben/a
  • –Tool poisoningtools not inspected (local package is not executed); not countedn/a
  • Auth qualitylocal package, no credentials required12/15
  • Maintenancelast push 39 days ago12/15
  • Maintainer identityregistry namespace matches repository owner; GitHub account older than a year8/10
Overall 88/100. Components that don't apply are left out of the denominator. Any critical finding is an F.RubricAppeal a findingJSON

How Tensorcad compares

Against every graded AI and agents MCP server, from our own scans and uptime checks. Updated after each scan run.

Safety score
88/100
higher than 80% of 4,460 AI and agents servers
Tools
19
the typical AI and agents server has 6
Tools that change things
21%
typical: 0%. Fewer write tools means less an agent can break
Downloads a week
149
more than 71% of 1,538 AI and agents servers with a package

What its tools let an agent do: 15 read · 4 change

Release history: 9 versions in the registry since 2026-09-23, latest 2026-09-26, typically several a day.

Install Tensorcad in Claude Code, Cursor or VS Code

Runs npx -y @tensor-cad/mcp on your machine. Read the scan report first; the gateway never runs local packages.

claude mcp add tensorcad -- npx -y @tensor-cad/mcp
Add to Cursor

What the publisher says

From the Tensorcad repository's README, as published. We do not edit it. Read it on GitHub

TensorCAD

Schematic capture for neural network architectures. Draw the model, get the numbers, generate the PyTorch.

Open the editor · Quick start · What it does · Correctness · For agents · Docs · Contributing · Roadmap · License

TensorCAD treats a neural network the way an EDA tool treats a circuit. Blocks are symbols with typed pins. Tensors are nets. Shapes are checked by a real algebra rather than by running the thing. A design-rule check tells you the model will not fit on your GPUs before you rent them. And the drawing is not a picture of the model — it is the model, and PyTorch falls out of it.

It started as a tool for language models. It now draws vision transformers and convolutional classifiers too, because the same machinery turned out to work.

Llama-3-8B, one level open. Every net carries its shape; the readout is measured under the operating point at the top of it, and the title block in the corner says how the count compares with the published one.

Fit it on the cluster. Every split the cluster admits, priced and ordered by how little it asks of you. Pressing one applies it, and the whole readout follows. Sweep small, run big. The same design at several widths, with what to multiply the initialization and the learning rate by at each. Press a rung to open it.

The same model as volumes, a port of Brendan Bycroft's LLM visualisation (MIT). Blue is a weight, green an activation, and a ribbon is a tensor on its way somewhere.

Shortened. The full README is on GitHub.

Nothing above is checked by us. What we check is on the safety report.

Tensorcad: common questions

Is Tensorcad MCP server safe?
Yes, by our scan: it is graded A (88/100). Read the Tensorcad safety report
How do I install Tensorcad?
It runs on your machine. Copy the Claude Code, Cursor, VS Code or Claude Desktop config from the install section.
Does Tensorcad need an API key?
No secret keys are declared. It reads 1 setting from the environment.
Is Tensorcad maintained?
The last commit was 39 days ago (2026-09-26). The latest release is v0.1.21.
What can I use instead of Tensorcad?
Servers from other publishers that do the same job: Figma Context MCP server, Basic Memory MCP server and Agent-Native Design MCP server.
How does Tensorcad compare with other AI and agents MCP servers?
safety score 88/100 (higher than 80% of 4,460 AI and agents servers); tools 19 (the typical AI and agents server has 6); tools that change things 21% (typical: 0%. Fewer write tools means less an agent can break). Full comparison.

Alternatives to Tensorcad

Same job from other publishers: the closest match first, then the best rated.

  • Figma Context
    Give your coding agent access to your Figma data. Implement designs in any framework in one-shot.
    B
  • Basic Memory
    Local-first knowledge management with bi-directional LLM sync via Markdown files.
    B
  • Agent-Native Design
    Agent-Native design tool - create and edit visual designs with agent assistance
    A
  • IWE (iwe.md)
    Markdown knowledge base as agent memory. Runs against the notes directory it is started in.
    A
  • Memanto
    MCP server for Memanto - persistent semantic memory for any MCP-compatible agent
    A