agentleFS
Sign inStart free
h9-tech9-tec/arabic-ai-atlas/skills/arabic-ai-advisor/SKILL.md

arabic-ai-advisor

Skill43 starsChanged 29 days agoMITView on GitHub

Use when choosing an Arabic LLM, ASR, TTS, OCR, embedding model, dataset or tool — recommends ranked options from the Arabic AI Atlas with licenses, dialect coverage and on-device fit.

We split this skill into the rules it gives an agent

6 rules. About 0 are common, 4 are distinctive, and 2 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈4

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 2

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Very specific:

    Call the MCP tool recommend with the task and the type the user wants (llm, asr, tts, ocr, embedding for a model.

    line 18
  2. Very specific:

    If the MCP tools are unavailable, read ${CLAUDE_PLUGIN_ROOT}/dist/atlas.json and filter entries by tasks.

    line 19
  3. Distinctive:

    If nothing fits, say: "Not in the atlas: I found no entry matching this request." Do not guess.

    line 27
  4. Distinctive:

    Never invent entries; only report what the atlas returns.

    line 31
  5. Distinctive:

    Prefer open licenses when the user does not say.

    line 32
  6. Distinctive:

    If two options tie, prefer the one updated more recently.

    line 34

≈ “Common” is an estimate: at least 1 in 5 of the 308,232 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Arabic AI Advisor

Recommends Arabic AI models, datasets and tools from the Arabic AI Atlas, a curated catalogue with Hugging Face download metrics.

When to use

  • The user needs an Arabic (or dialect-specific) LLM, ASR, TTS, OCR or embedding model.
  • The user asks which Arabic dataset, benchmark or tool fits a task.
  • The user has constraints such as a license, a dialect, or running on-device.

Procedure

  1. Call the MCP tool recommend with the task and the type the user wants (llm, asr, tts, ocr, embedding for a model; dataset, benchmark or tool otherwise), plus dialect, on_device and license_filter when the user gave them. Example: recommend(task="tts", type="tts", on_device=true, license_filter="open").
  2. If the MCP tools are unavailable, read ${CLAUDE_PLUGIN_ROOT}/dist/atlas.json and filter entries by tasks.
  3. For breadth, or when recommend returns too little, call search (and get for one entry by id). To trace what a model was built from or what was built on it, call lineage(id=...); entries carry an optional base_model list (HF ids, atlas ids or from-scratch).
  4. Format the answer as described below.

Answer format

Give 3 ranked options. For each: name, org, license, dialects, size, a one-line why, and a link. Then state the atlas generated_at date.

If nothing fits, say: "Not in the atlas: I found no entry matching this request." Do not guess.

Rules

  • Never invent entries; only report what the atlas returns.
  • Prefer open licenses when the user does not say.
  • Say when metrics (downloads, size, license) are missing.
  • If two options tie, prefer the one updated more recently.

Valid values

ParameterValues
typellm, asr, tts, ocr, embedding, dataset, benchmark, tool, agent-skill, org, paper
countrySA, AE, EG, LB, QA, JO, MA, TN, KW, OM, BH, DZ, LY, SD, IQ, SY, YE, PS, MR, SO, DJ, KM, INTL
modalitytext, speech, vision, multimodal, none
dialectmsa, egy, gulf, lev, magh, iraqi, sudanese, yemeni, classical, mixed
base_modellist of HF ids, atlas ids, or from-scratch (model types only)
task (examples)chat, tts, asr, ocr, embedding, translation

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

Check your setup, free