modular
Safety scan
A static text scan, not a guarantee. Select the badge to see what was scanned and every finding.
Check it deeper on your machine
It has changed since its scan on 11 Nov 2026.
As a Claude Code plugin, install it and then ask Claude Code:
/plugin marketplace add keithmackay/mackayi/plugin install scanskillsafety@mackayiDo a deep scan of https://github.com/modular/skills with scanskillsafety
Or from a terminal, clone it and run it:
git clone https://github.com/keithmackay/scanskillsafety-skill ~/.claude/skills/scanskillsafetynode ~/.claude/skills/scanskillsafety/cli.cjs --deep 'https://github.com/modular/skills'
It checks the exact commit: every file, its packages against OSV (api.osv.dev), how it compares with this page's scan (findsafeskills.com receives the repository's address), and, if you agree, Claude reading every file with your own Claude subscription. It never runs the code.
Badge for your README
Markdown
[](/listings/modular--skills)HTML
<a href="/listings/modular--skills"><img src="/api/listings/modular--skills/badge.svg" alt="findsafeskills safety rating" /></a>Install
Install instructions from the README
As written by the repo's author. findsafeskills has not verified these steps.
Pick the instructions for your agent:
Claude Code
/plugin marketplace add modular/skills /plugin install skills@modular
For a subset, install `max-skills@modular` to work with models on MAX, or `mojo-skills@modular` to write Mojo.
Codex and other agents
With [Node.js installed](https://nodejs.org/en/download):
npx skills add modular/skills
This installs the skills into the right location for your agent.
Update them later:
npx skills update
Install one skill at a time:
npx skills add modular/skills --skill mojo-syntax
To install by hand, clone [the repository](https://github.com/modular/skills) and copy or symlink the individual skill directories into your agent's skills directory (for Claude Code, `~/.claude/skills/`).
What was scanned
The files the safety scan read, as written by the repository's author and shown as plain text. Lines with a finding are marked.
.claude-plugin/marketplace.json · Manifest · 60 lines
View this file on GitHub (read at the repository's current version: the scan did not record which commit it read)
This file has changed since it was scanned, so this is not exactly the text that was scanned, and no line is marked. The findings are listed with the safety scan above.
- {
- "name": "modular",
- "owner": {
- "name": "Modular",
- "email": "hello@modular.com"
- },
- "metadata": {
- "description": "Official Modular agent skills for MAX and Mojo",
- "version": "26.5"
- },
- "plugins": [
- {
- "name": "skills",
- "description": "All official Modular skills for MAX and Mojo",
- "source": "./",
- "strict": false,
- "skills": [
- "./new-modular-project",
- "./mojo-syntax",
- "./mojo-gpu-fundamentals",
- "./mojo-python-interop",
- "./import-model",
- "./migrate-max-v2-to-v3",
- "./debug-model",
- "./serve-model",
- "./benchmark-model",
- "./profile-model",
- "./eval-model"
- ]
- },
- {
- "name": "max-skills",
- "description": "Modular skills for importing, serving, and measuring models on MAX",
- "source": "./",
- "strict": false,
- "skills": [
- "./new-modular-project",
- "./import-model",
- "./migrate-max-v2-to-v3",
- "./debug-model",
- "./serve-model",
- "./benchmark-model",
- "./profile-model",
- "./eval-model"
- ]
- },
- {
- "name": "mojo-skills",
- "description": "Modular skills for writing Mojo, including GPU and Python interop",
- "source": "./",
- "strict": false,
- "skills": [
- "./new-modular-project",
- "./mojo-syntax",
- "./mojo-gpu-fundamentals",
- "./mojo-python-interop"
- ]
- }
- ]
- }
README.md · README · 128 lines
View this file on GitHub (read at the repository's current version: the scan did not record which commit it read)
This file has changed since it was scanned, so this is not exactly the text that was scanned, and no line is marked. The findings are listed with the safety scan above.
- [](https://skills.sh/modular/skills)
- # Modular skills
- These are the official AI agent skills from [Modular](https://www.modular.com/)
- for working with the Modular Platform, including MAX and Mojo. They follow the
- [Agent Skills Standard](https://agentskills.io/specification). Any AI coding
- agent can use them to write Mojo, or to import, serve, and measure models on
- MAX.
- ## Install
- Pick the instructions for your agent:
- ### Claude Code
- ```text
- /plugin marketplace add modular/skills
- /plugin install skills@modular
- ```
- For a subset, install `max-skills@modular` to work with models on MAX, or
- `mojo-skills@modular` to write Mojo.
- ### Codex and other agents
- With [Node.js installed](https://nodejs.org/en/download):
- ```bash
- npx skills add modular/skills
- ```
- This installs the skills into the right location for your agent.
- Update them later:
- ```bash
- npx skills update
- ```
- Install one skill at a time:
- ```bash
- npx skills add modular/skills --skill mojo-syntax
- ```
- To install by hand, clone
- [the repository](https://github.com/modular/skills) and copy or symlink the
- individual skill directories into your agent's skills directory (for Claude
- Code, `~/.claude/skills/`).
- ## Mojo and project skills
- Skills that cover writing Mojo and setting up a project:
- - [`/new-modular-project`](new-modular-project/SKILL.md): Creates a new Mojo or
- MAX project, setting up the `pixi` or `uv` environment for you.
- - [`/mojo-syntax`](mojo-syntax/SKILL.md): Corrects pretrained assumptions so
- your agent writes modern Mojo. Use it whenever an agent writes Mojo.
- - [`/mojo-gpu-fundamentals`](mojo-gpu-fundamentals/SKILL.md): Adds the patterns
- for programming NVIDIA, AMD, and Apple silicon GPUs in Mojo. Pairs with
- `/mojo-syntax`.
- - [`/mojo-python-interop`](mojo-python-interop/SKILL.md): Handles Mojo calling
- Python and Python calling Mojo, including building Python extension modules.
- Pairs with `/mojo-syntax`.
- ## Model lifecycle
- Skills that take a model from a Hugging Face checkpoint to a deployment you've
- verified and measured. They hand off to each other, and each one names the state
- it expects the model to be in:
- - [`/import-model`](import-model/SKILL.md): Imports a new model architecture
- into MAX from a Hugging Face model ID as a ModuleV3 model
- (`max.nn`), scaffolding from a similar registered architecture
- and verifying outputs match. Hands off to `/debug-model` when
- the server runs but the text is wrong.
- - [`/serve-model`](serve-model/SKILL.md): Takes you from no environment to a
- running OpenAI-compatible endpoint with `max serve`, choosing the flags a
- model needs: `--devices`, `--quantization-encoding`, `--max-length`,
- `--task`, and `--trust-remote-code`. Use
- `--custom-architectures` for an architecture you ported with
- `/import-model`.
- - [`/migrate-max-v2-to-v3`](migrate-max-v2-to-v3/SKILL.md): Ports a model
- written against the V2 graph API (`max.nn.legacy`, `TensorValue`, explicit `Graph`)
- to ModuleV3 (`max.nn`), leaving the V2 code untouched and
- verifying greedy outputs match.
- - [`/debug-model`](debug-model/SKILL.md): Takes over once a model loads and
- generates tokens but the output is wrong. Builds tensor-dump comparators and
- bisects serve versus pipeline. For crashes on load, use `/import-model`.
- - [`/benchmark-model`](benchmark-model/SKILL.md): Drives load against an
- endpoint you started with `/serve-model`. Reports throughput and latency
- (TTFT, TPOT, inter-token latency), plus GPU utilization when it runs on the
- same NVIDIA host as the server. Hand off to `/profile-model` when the numbers
- show a bottleneck.
- - [`/profile-model`](profile-model/SKILL.md): Finds where inference time goes
- and whether the GPU is saturated. Works cheapest-first: a utilization
- snapshot, a kernel breakdown with `nsys` or `rocprofv3`, then an `ncu` deep
- dive on a single kernel when one dominates.
- - [`/eval-model`](eval-model/SKILL.md): Measures task accuracy on standard
- benchmarks: GSM8K, MMLU, HellaSwag, ARC, AIME, GPQA, TruthfulQA, WinoGrande,
- and BABILong. Distinguishes serving failures from wrong model answers. Ask
- for it by name. It won't trigger on its own.
- ## Examples
- Once you install these skills, you can use them for many common tasks.
- Examples include:
- ### Start a new Mojo project
- ```text
- I'd like to create a new Mojo project named "nvfp4-for-metal".
- ```
- ### Translate CUDA C++ code to Mojo
- ```text
- A CUDA kernel is present in `../example`, please create a new Mojo project that implements that same kernel.
- ```
- For several of these skills, your AI agent may prompt you for more information
- to clarify your objectives and to make sure it uses the right tools and
- patterns.
- ## License
- Apache 2.0. See the [LICENSE](./LICENSE) file for details.
Ratings and comments
Sign in to rate this skill.
Sign inFree to use. If it’s helped, pass it on.