Skip to content
Back to search

infra-skills

Open, portable agent skills for AI infrastructure engineering.

Safety scan

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

Checked today
Badge for your README
findsafeskills safety rating

Markdown

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

HTML

<a href="/listings/open-infra-skills--infra-skills"><img src="/api/listings/open-infra-skills--infra-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.

The repository follows the open Agent Skills directory convention. Use the Skills CLI to inspect or install it:

# List all skills in the repository.
npx skills add open-infra-skills/infra-skills --list

# Install one skill interactively.
npx skills add open-infra-skills/infra-skills --skill optimize-musa-training

# Install one skill globally for Codex without prompts.
npx skills add open-infra-skills/infra-skills \
  --skill optimize-musa-training \
  --global \
  --agent codex \
  --yes

You can also clone the repository and point a compatible agent at a skill directory directly. Each skill keeps its instructions and optional scripts, references, and agent metadata together.

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.

README.md · README · 115 lines

View this file on GitHub (read at the repository's current version: the scan did not record which commit it read)

  1. # Open Infra Skills
  2. Open, portable agent skills for building, operating, profiling, and optimizing AI infrastructure.
  3. This repository is intended to grow into a comprehensive, vendor-aware library for the full AI systems lifecycle: accelerator bring-up, distributed training, inference serving, kernels and compilers, data pipelines, observability, benchmarking, and cluster operations.
  4. ## Why This Repository Exists
  5. AI infrastructure work is full of environment-specific details, expensive experiments, and performance claims that are easy to mismeasure. Open Infra Skills packages repeatable engineering workflows as self-contained agent skills so that an assistant can:
  6. - establish a reproducible baseline before changing a system;
  7. - choose the right diagnostic layer, from application timing to system traces and kernel counters;
  8. - preserve numerical behavior while optimizing performance;
  9. - distinguish measured evidence from hypotheses;
  10. - work across local machines, SSH hosts, schedulers, containers, and clouds without assuming one client operating system.
  11. ## Skill Catalog
  12. | Skill | Category | What it helps with |
  13. | --- | --- | --- |
  14. | [`optimize-musa-training`](skills/accelerators/optimize-musa-training/SKILL.md) | Accelerators | Profile, debug, and optimize AI training on Moore Threads MUSA GPUs, including MFU/HFU, Moore Perf, FA2, `torch.compile`, FSDP, MCCL, and kernel-level analysis. |
  15. The machine-readable catalog is available in [`catalog.json`](catalog.json).
  16. ## Install
  17. The repository follows the open Agent Skills directory convention. Use the Skills CLI to inspect or install it:
  18. ```bash
  19. # List all skills in the repository.
  20. npx skills add open-infra-skills/infra-skills --list
  21. # Install one skill interactively.
  22. npx skills add open-infra-skills/infra-skills --skill optimize-musa-training
  23. # Install one skill globally for Codex without prompts.
  24. npx skills add open-infra-skills/infra-skills \
  25. --skill optimize-musa-training \
  26. --global \
  27. --agent codex \
  28. --yes
  29. ```
  30. You can also clone the repository and point a compatible agent at a skill directory directly. Each skill keeps its instructions and optional scripts, references, and agent metadata together.
  31. ## Repository Layout
  32. ```text
  33. infra-skills/
  34. |-- skills/
  35. | `-- <category>/
  36. | `-- <skill-name>/
  37. | |-- SKILL.md
  38. | |-- agents/ # Optional agent-facing metadata
  39. | |-- references/ # Optional progressive-disclosure docs
  40. | `-- scripts/ # Optional deterministic helpers
  41. |-- scripts/ # Repository validation and catalog tooling
  42. |-- .github/ # CI, issue forms, and PR guidance
  43. |-- catalog.json # Generated skill index
  44. `-- README.md
  45. ```
  46. The category layer keeps a large library navigable while remaining compatible with tools that discover `skills/<category>/<name>/SKILL.md`.
  47. ## Scope
  48. Planned categories include:
  49. - `accelerators`: vendor runtimes, device bring-up, profiling, and architecture-aware optimization;
  50. - `distributed-training`: DDP, FSDP, tensor/pipeline parallelism, collectives, and failure diagnosis;
  51. - `inference-serving`: throughput, latency, batching, caching, quantization, and serving engines;
  52. - `compilers-kernels`: graph compilers, custom operators, Triton/CUDA/MUSA kernels, and numerical validation;
  53. - `data-storage`: preprocessing, dataloaders, object storage, checkpoint I/O, and data movement;
  54. - `observability-benchmarking`: MFU/HFU, traces, rooflines, reproducible benchmarks, and cost analysis;
  55. - `cluster-operations`: schedulers, containers, networking, health checks, and incident workflows.
  56. This scope is a direction, not a requirement to create empty scaffolding. Categories are added when a maintained skill exists.
  57. ## Design Principles
  58. 1. **Evidence first.** Measure before optimizing, and state exactly what a metric includes.
  59. 2. **Correctness is a gate.** A speedup is retained only after numerical and behavioral checks pass.
  60. 3. **Portable orchestration.** Skills describe target-system behavior without requiring a particular laptop OS, IDE, shell transport, or scheduler.
  61. 4. **Progressive disclosure.** Keep the main `SKILL.md` focused and load deeper references only when needed.
  62. 5. **Reproducible experiments.** Record source revision, environment, command, workload shape, warmup policy, and raw evidence.
  63. 6. **Secure by default.** Never publish credentials, private hosts, proprietary data paths, or internal registry details.
  64. 7. **Honest negative results.** Document failed or neutral experiments when they prevent others from repeating expensive work.
  65. ## Quality Bar
  66. Every skill must:
  67. - have valid `name` and `description` frontmatter;
  68. - use a lowercase kebab-case directory matching its skill name;
  69. - contain actionable workflows rather than a loose collection of tips;
  70. - separate facts, measurements, inferences, and recommendations;
  71. - include verification steps and failure boundaries;
  72. - avoid private infrastructure details and embedded secrets;
  73. - pass repository validation and keep generated catalog data current.
  74. Run the checks locally with:
  75. ```bash
  76. npm test
  77. ```
  78. See [`CONTRIBUTING.md`](CONTRIBUTING.md) for the authoring and review workflow.
  79. ## Project Status
  80. Open Infra Skills is in early development. The first skill focuses on MUSA training optimization; the repository structure and validation pipeline are designed to support a much broader AI infrastructure collection without turning every skill into one giant prompt.
  81. ## License
  82. Licensed under the [Apache License 2.0](LICENSE).

Ratings and comments

Sign in to rate this skill.

Sign in

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