Skip to content
Back to search

modular

SkillGitHubIncomplete

Safety scan

A static text scan, not a guarantee. Select the badge to see what was scanned and every finding.

Checked today

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
findsafeskills safety rating

Markdown

[![findsafeskills safety rating](/api/listings/modular--skills/badge.svg)](/listings/modular--skills)

HTML

<a href="/listings/modular--skills"><img src="/api/listings/modular--skills/badge.svg" alt="findsafeskills safety rating" /></a>

Install

i

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.

  1. {
  2. "name": "modular",
  3. "owner": {
  4. "name": "Modular",
  5. "email": "hello@modular.com"
  6. },
  7. "metadata": {
  8. "description": "Official Modular agent skills for MAX and Mojo",
  9. "version": "26.5"
  10. },
  11. "plugins": [
  12. {
  13. "name": "skills",
  14. "description": "All official Modular skills for MAX and Mojo",
  15. "source": "./",
  16. "strict": false,
  17. "skills": [
  18. "./new-modular-project",
  19. "./mojo-syntax",
  20. "./mojo-gpu-fundamentals",
  21. "./mojo-python-interop",
  22. "./import-model",
  23. "./migrate-max-v2-to-v3",
  24. "./debug-model",
  25. "./serve-model",
  26. "./benchmark-model",
  27. "./profile-model",
  28. "./eval-model"
  29. ]
  30. },
  31. {
  32. "name": "max-skills",
  33. "description": "Modular skills for importing, serving, and measuring models on MAX",
  34. "source": "./",
  35. "strict": false,
  36. "skills": [
  37. "./new-modular-project",
  38. "./import-model",
  39. "./migrate-max-v2-to-v3",
  40. "./debug-model",
  41. "./serve-model",
  42. "./benchmark-model",
  43. "./profile-model",
  44. "./eval-model"
  45. ]
  46. },
  47. {
  48. "name": "mojo-skills",
  49. "description": "Modular skills for writing Mojo, including GPU and Python interop",
  50. "source": "./",
  51. "strict": false,
  52. "skills": [
  53. "./new-modular-project",
  54. "./mojo-syntax",
  55. "./mojo-gpu-fundamentals",
  56. "./mojo-python-interop"
  57. ]
  58. }
  59. ]
  60. }
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.

  1. [![skills.sh](https://skills.sh/b/modular/skills)](https://skills.sh/modular/skills)
  2. # Modular skills
  3. These are the official AI agent skills from [Modular](https://www.modular.com/)
  4. for working with the Modular Platform, including MAX and Mojo. They follow the
  5. [Agent Skills Standard](https://agentskills.io/specification). Any AI coding
  6. agent can use them to write Mojo, or to import, serve, and measure models on
  7. MAX.
  8. ## Install
  9. Pick the instructions for your agent:
  10. ### Claude Code
  11. ```text
  12. /plugin marketplace add modular/skills
  13. /plugin install skills@modular
  14. ```
  15. For a subset, install `max-skills@modular` to work with models on MAX, or
  16. `mojo-skills@modular` to write Mojo.
  17. ### Codex and other agents
  18. With [Node.js installed](https://nodejs.org/en/download):
  19. ```bash
  20. npx skills add modular/skills
  21. ```
  22. This installs the skills into the right location for your agent.
  23. Update them later:
  24. ```bash
  25. npx skills update
  26. ```
  27. Install one skill at a time:
  28. ```bash
  29. npx skills add modular/skills --skill mojo-syntax
  30. ```
  31. To install by hand, clone
  32. [the repository](https://github.com/modular/skills) and copy or symlink the
  33. individual skill directories into your agent's skills directory (for Claude
  34. Code, `~/.claude/skills/`).
  35. ## Mojo and project skills
  36. Skills that cover writing Mojo and setting up a project:
  37. - [`/new-modular-project`](new-modular-project/SKILL.md): Creates a new Mojo or
  38. MAX project, setting up the `pixi` or `uv` environment for you.
  39. - [`/mojo-syntax`](mojo-syntax/SKILL.md): Corrects pretrained assumptions so
  40. your agent writes modern Mojo. Use it whenever an agent writes Mojo.
  41. - [`/mojo-gpu-fundamentals`](mojo-gpu-fundamentals/SKILL.md): Adds the patterns
  42. for programming NVIDIA, AMD, and Apple silicon GPUs in Mojo. Pairs with
  43. `/mojo-syntax`.
  44. - [`/mojo-python-interop`](mojo-python-interop/SKILL.md): Handles Mojo calling
  45. Python and Python calling Mojo, including building Python extension modules.
  46. Pairs with `/mojo-syntax`.
  47. ## Model lifecycle
  48. Skills that take a model from a Hugging Face checkpoint to a deployment you've
  49. verified and measured. They hand off to each other, and each one names the state
  50. it expects the model to be in:
  51. - [`/import-model`](import-model/SKILL.md): Imports a new model architecture
  52. into MAX from a Hugging Face model ID as a ModuleV3 model
  53. (`max.nn`), scaffolding from a similar registered architecture
  54. and verifying outputs match. Hands off to `/debug-model` when
  55. the server runs but the text is wrong.
  56. - [`/serve-model`](serve-model/SKILL.md): Takes you from no environment to a
  57. running OpenAI-compatible endpoint with `max serve`, choosing the flags a
  58. model needs: `--devices`, `--quantization-encoding`, `--max-length`,
  59. `--task`, and `--trust-remote-code`. Use
  60. `--custom-architectures` for an architecture you ported with
  61. `/import-model`.
  62. - [`/migrate-max-v2-to-v3`](migrate-max-v2-to-v3/SKILL.md): Ports a model
  63. written against the V2 graph API (`max.nn.legacy`, `TensorValue`, explicit `Graph`)
  64. to ModuleV3 (`max.nn`), leaving the V2 code untouched and
  65. verifying greedy outputs match.
  66. - [`/debug-model`](debug-model/SKILL.md): Takes over once a model loads and
  67. generates tokens but the output is wrong. Builds tensor-dump comparators and
  68. bisects serve versus pipeline. For crashes on load, use `/import-model`.
  69. - [`/benchmark-model`](benchmark-model/SKILL.md): Drives load against an
  70. endpoint you started with `/serve-model`. Reports throughput and latency
  71. (TTFT, TPOT, inter-token latency), plus GPU utilization when it runs on the
  72. same NVIDIA host as the server. Hand off to `/profile-model` when the numbers
  73. show a bottleneck.
  74. - [`/profile-model`](profile-model/SKILL.md): Finds where inference time goes
  75. and whether the GPU is saturated. Works cheapest-first: a utilization
  76. snapshot, a kernel breakdown with `nsys` or `rocprofv3`, then an `ncu` deep
  77. dive on a single kernel when one dominates.
  78. - [`/eval-model`](eval-model/SKILL.md): Measures task accuracy on standard
  79. benchmarks: GSM8K, MMLU, HellaSwag, ARC, AIME, GPQA, TruthfulQA, WinoGrande,
  80. and BABILong. Distinguishes serving failures from wrong model answers. Ask
  81. for it by name. It won't trigger on its own.
  82. ## Examples
  83. Once you install these skills, you can use them for many common tasks.
  84. Examples include:
  85. ### Start a new Mojo project
  86. ```text
  87. I'd like to create a new Mojo project named "nvfp4-for-metal".
  88. ```
  89. ### Translate CUDA C++ code to Mojo
  90. ```text
  91. A CUDA kernel is present in `../example`, please create a new Mojo project that implements that same kernel.
  92. ```
  93. For several of these skills, your AI agent may prompt you for more information
  94. to clarify your objectives and to make sure it uses the right tools and
  95. patterns.
  96. ## License
  97. Apache 2.0. See the [LICENSE](./LICENSE) file for details.

Ratings and comments

Sign in to rate this skill.

Sign in

Free to use. If it’s helped, pass it on.