Agent skill

Primus Projection

by AMD-AGI in AMD-AGI/Primus

Opinionated guide for using Primus Projection to choose parallelism (TP/PP/EP/CP/DP) and pipeline schedules, validate memory fit on target nodes, reason about communication collectives, and explore…

Custom licenceAuto-check passed

Install Primus Projection

skills CLI
$ npx skills add AMD-AGI/Primus --skill primus-projection -a claude-code

Project install by default; add -g for ~/.claude/skills/.

GitHub CLI
$ gh skill install AMD-AGI/Primus primus-projection --agent claude-code

Project scope by default; add --scope user for a personal install. Needs GitHub CLI 2.90.0 or later (public preview).

Manual copy
$ git clone --depth 1 https://github.com/AMD-AGI/Primus.git skills-src && mkdir -p .claude/skills && cp -r skills-src/skills/primus-projection .claude/skills/primus-projection && rm -rf skills-src

Use ~/.claude/skills/ instead of .claude/skills for a personal install. The folder must contain SKILL.md.

Claude Code skills documentation · loads skills from .claude/skills/

Facts

Skill name
primus-projection
GitHub stars
131
Token cost
~5.3k tokens
SKILL.md length
1,973 words
Files
1
Skills in repo
9
Repo updated
First seen
Licence
Custom licence

At a glance

Opinionated guide for using Primus Projection to choose parallelism (TP/PP/EP/CP/DP) and pipeline schedules, validate memory fit on target nodes, reason about communication collectives, and explore…

  • Works in 3 steps: DeepEP — single largest gain:… → SyncFree stages — progressive fusion:… → EP size — compare: --target-ep-size 4 vs…
  • The user asks how to pick a parallelism strategy
  • SKILL.md covers Purpose, Scope, Training Backends and Projection Advisor (What users…, plus 3 more sections
  • Calls bash

What it does

Primus Projection is an agent skill from AMD-AGI/Primus. Opinionated guide for using Primus Projection to choose parallelism (TP/PP/EP/CP/DP) and pipeline schedules, validate memory fit on target nodes, reason about communication collectives, and explore optimization trade-offs with minimal compute. Use when the user asks how to pick a parallelism strategy or pipeline schedule, whether a model fits in memory, which optimizations such as DeepEP, SyncFree, zero-bubble, FP8, recomputation, or FSDP2 matter most, or how to run primus projection commands. Read-only planning…

Its SKILL.md is about 5.3k tokens, which your agent loads only when the skill is triggered. It is a single SKILL.md file with no bundled scripts.

It works with NVIDIA AI Platform. The repository describes itself as: A flexible and high-performance training framework designed for large-scale foundation model training on AMD GPUs.

When your agent uses it

  • The user asks how to pick a parallelism strategy
  • Pipeline schedule
  • Whether a model fits in memory
  • Which optimizations such as DeepEP

Example prompts

  • “/primus-projection”

Workflow steps

3 steps, taken from the first numbered list in SKILL.md.

  1. DeepEP — single largest gain: --enable-deepep
  2. SyncFree stages — progressive fusion: --sync-free-stage 2, then 3
  3. EP size — compare: --target-ep-size 4 vs --target-ep-size 8

What it can do on your machine

Read from SKILL.md and the folder at commit 0f99649. It shows what the files ask for, not the result of running them.

  • Tool permissions

    Pre-approves nothing: there is no allowed-tools line, so your agent's usual permission prompts apply.

    From allowed-tools in the SKILL.md frontmatter.

  • Runs code

    Shell commands in SKILL.md call:

    • bash

    From the folder's file list and the shell code blocks in SKILL.md.

  • Network

    Links to these hosts (documentation or services it may open):

    • rocm.blogs.amd.com

    From URLs in SKILL.md, links to its own repository left out.

  • Credentials

    Names no API keys, tokens, secrets or passwords.

    From names ending in _API_KEY, _TOKEN, _SECRET, _KEY or _PASSWORD in SKILL.md.

Context cost

Primus Projection loads about 5.3k tokens when it runs. Until then it costs about 147 tokens; SKILL.md has 1,973 words of instructions outside code blocks.

Always · name and description, kept in context so the agent knows when to use it
~147
When it runs · the whole SKILL.md, loaded when a task matches
~5.3k

Estimates: characters ÷ 4, the usual rule of thumb; real counts depend on the model's tokenizer. Scripts and assets cost tokens only if the agent reads them.

Safety

Auto-check passed

The automated check found no risky patterns in SKILL.md.

Automated static check — not a guarantee. Review scripts before installing. It scans the text of SKILL.md for risky patterns (piping downloads into a shell, reading credential files, hidden Unicode, destructive commands); files beside SKILL.md are not scanned.

SKILL.md

Its licence (Custom licence) doesn't allow us to republish the file, so here is its outline and opening line. It has 1,973 words (~5,250 tokens).

“Provide a quick, opinionated guide for using Primus Projection to:”

— opening of SKILL.md by AMD-AGI, Custom licence
name
primus-projection

Read the full SKILL.md on GitHub

Files

Just SKILL.md in skills/primus-projection of AMD-AGI/Primus.

Open the folder on GitHubat commit 0f99649

Compare with similar skills

Primus Projection next to the 5 skills that share the most tags, products or categories with it. Stars are the repository's; “used in” counts other GitHub owners with a copy.

Primus Projection compared with similar skills
SkillStarsUsed inTokensAuto-checkLicenceRepo updated
Primus Projection this skillAMD-AGI/Primus131—~5.3kAutomated safety check: PassCustom licence
Skill InspectorNVIDIA/SkillSpector20k1 repos~1.8kAutomated safety check: PassApache-2.0
LLM Torch Profiler Analysissgl-project/sglang37k2 repos~6.4kAutomated safety check: PassApache-2.0
Embeddings via 9Routerdecolua/9router30k—~604Automated safety check: PassMIT
NEAR AI Cloud Private Inferenceinternet-court/internet-court-skill6.4k2 repos~1.3kAutomated safety check: PassCustom licence
Nemoclaw Maintainer Normalize Title TagsNVIDIA/NemoClaw23k—~693Automated safety check: PassApache-2.0

Similar skills

  • Skill Inspector

    NVIDIA/SkillSpector

    Official

    Decides whether an agent skill is safe to install by combining a SkillSpector static scan with the agent's own source review, ending in APPROVE, CAUTION or REJECT.

    20k GitHub starsUsed in 1 repo~1.8k tokens
    SecurityAuto-check passed
  • LLM Torch Profiler Analysis

    sgl-project/sglang

    Unified LLM torch-profiler triage skill for sglang, vllm, TensorRT-LLM, and TokenSpeed.

    37k GitHub starsUsed in 2 repos~6.4k tokens
    DevelopmentAuto-check passed
  • Embeddings via 9Router

    decolua/9router

    Generates vector embeddings through the 9Router /v1/embeddings endpoint, using models from providers such as OpenAI, Gemini, Mistral and Voyage for RAG and semantic search.

    30k GitHub stars~604 tokensUpdated 7 days ago
    AI & LLM EngineeringAuto-check passed
  • NEAR AI Cloud Private Inference

    internet-court/internet-court-skill

    Shows how to call NEAR AI Cloud through an OpenAI-compatible API and verify that inference ran in a TEE, using attestation checks and signed chat responses.

    6.4k GitHub starsUsed in 2 repos~1.3k tokens
    AI & LLM EngineeringAuto-check passed
  • Remove bracketed NemoClaw tags from GitHub issue and PR titles.

    23k GitHub stars~693 tokensUpdated today
    Marketing & SEOAuto-check passed
  • Audit and implement a NemoClaw dependency version upgrade, including Hermes and base images.

    23k GitHub stars~1.4k tokensUpdated today
    DevelopmentAuto-check passed

More from AMD-AGI/Primus

All 9 skills in this repo
  • Spur Cluster Status

    AMD-AGI/Primus

    Inspect the current Spur (AMD SLURM-compatible) cluster node-allocation state and produce a Markdown report (optionally rendered to a standalone HTML page) covering the caller's account/QoS…

    131 GitHub stars~2.7k tokensUpdated today
    Auto-check passed
  • Backend Gap Report

    AMD-AGI/Primus

    Compare a Primus backend against an upstream repository or reference, verify git state, dependencies, directory changes, and integration coupling, then generate comparison reports, dashboard…

    131 GitHub stars~3.4k tokensUpdated today
    Auto-check passed
  • Inventory and explain the patch (monkey-patch) optimizations Primus layers over upstream training backends such as Megatron-LM, TorchTitan, and MaxText, including their version compatibility…

    131 GitHub stars~3.5k tokensUpdated today
    Auto-check passed
  • Convergence Test

    AMD-AGI/Primus

    Run, monitor, stop and report Primus convergence tests -- training a model on a real corpus and checking that the loss curve is healthy -- from a plain-language request such as "run convergence test…

    131 GitHub stars~2.1k tokensUpdated today
    Auto-check passed
  • Release Docs Update

    AMD-AGI/Primus

    Update the Primus documentation for a monthly training-image release — bump version references, realign the bare-metal install scripts with the release Dockerfiles, record the published image stacks…

    131 GitHub stars~2.9k tokensUpdated today
    Auto-check passed
  • Validate SLURM cluster nodes by running actual training jobs in groups.

    131 GitHub stars~3.7k tokensUpdated today
    Auto-check passed

Questions about Primus Projection

What does Primus Projection do?

Opinionated guide for using Primus Projection to choose parallelism (TP/PP/EP/CP/DP) and pipeline schedules, validate memory fit on target nodes, reason about communication collectives, and explore…. Primus Projection is an agent skill from AMD-AGI/Primus. Opinionated guide for using Primus Projection to choose parallelism (TP/PP/EP/CP/DP) and pipeline schedules, validate memory fit on target nodes, reason about communication collectives, and explore optimization trade-offs with minimal compute.

When should I use Primus Projection?

Primus Projection fits situations like: the user asks how to pick a parallelism strategy; pipeline schedule; whether a model fits in memory; which optimizations such as DeepEP.

How do I install Primus Projection in Claude Code?

Run `npx skills add AMD-AGI/Primus --skill primus-projection -a claude-code`. Or copy the skill folder (skills/primus-projection in AMD-AGI/Primus) into .claude/skills/primus-projection in your project. Claude Code loads it when a task matches its description.

How do I install Primus Projection in Codex?

Run `npx skills add AMD-AGI/Primus --skill primus-projection -a codex`. Or copy the skill folder (skills/primus-projection in AMD-AGI/Primus) into .agents/skills/primus-projection in your project. Codex loads it when a task matches its description.

Can I use Primus Projection in Cursor, Gemini CLI or GitHub Copilot?

Cursor, Gemini CLI, GitHub Copilot and OpenCode also load SKILL.md folders. With the skills CLI, run `npx skills add AMD-AGI/Primus --skill primus-projection -a cursor` (or -a gemini-cli, github-copilot or opencode for the others). To copy it by hand, put the folder in .cursor/skills/primus-projection, .gemini/skills/primus-projection, .github/skills/primus-projection and .opencode/skills/primus-projection in your project.

What does Primus Projection need to run?

Going by SKILL.md and its folder, Primus Projection needs the command-line tools its instructions call (bash).

Does Primus Projection access the network?

SKILL.md names 1 domain. As links in the text: rocm.blogs.amd.com. This is read from the text; nothing was executed.

Is Primus Projection safe to install?

Our automated static check of SKILL.md found no risky patterns, such as piping downloads into a shell, reading credential files or hidden Unicode. It is not a guarantee. Review the folder before installing.

What licence does Primus Projection use?

Primus Projection has a licence file (the repository's licence) that doesn't match a standard licence. Read it on GitHub before reusing the skill.

How many tokens does Primus Projection use?

About 5.3k tokens (SKILL.md is roughly 21k characters). Agents keep only the skill's name and description in context until a task matches; then they load SKILL.md in full.

What are the alternatives to Primus Projection?

Skills that share tags, products or a category with Primus Projection: Skill Inspector (NVIDIA/SkillSpector, 20k stars), LLM Torch Profiler Analysis (sgl-project/sglang, 37k stars), Embeddings via 9Router (decolua/9router, 30k stars) and NEAR AI Cloud Private Inference (internet-court/internet-court-skill, 6.4k stars). The comparison table on this page puts their stars, adoption, token cost, safety result and licence side by side.

Who maintains Primus Projection?

AMD-AGI (a GitHub organization) maintains it in AMD-AGI/Primus, which has 131 GitHub stars. The repository holds 9 skills in this directory. The repository was last updated on October 8, 2026.

Source: AMD-AGI/Primus on GitHub. Facts on this page come from the repository at the commit we read; the author's words are quoted as theirs.