Skill Title

One-paragraph overview: what this skill does and the value it adds. Keep the whole SKILL.md body under ~500 lines — move long detail into references/ (loaded on demand).

When to Use This Skill

Use this skill when the user wants to:

• <concrete task 1>
• <concrete task 2>

Does NOT Trigger

Scenario	Use Instead
<adjacent task>	alterlab-other-skill

Core Capabilities

1. <Capability>

Concrete, correct guidance with real APIs/commands. Cite real sources; never fabricate citations, benchmarks, or DOIs.

Resources

• references/<topic>.md — detailed reference loaded on demand
• scripts/<tool>.py — runnable helper (invoke as python scripts/<tool>.py)

<!-- AUTHORING CHECKLIST (see CONTRIBUTING.md → Skill Quality Standards): - name == this directory's name, lowercase-hyphen, no 'claude'/'anthropic' - description: third person, leads with what + "Use when", suite label LAST, <=1024 chars - body <500 lines; reference files exist; tools scoped in allowed-tools - add evals/evals.json on the docs/evals.schema.json shape (>=3 should_trigger + >=1 should_not_trigger assertion — see docs/evals.md) - validate: python scripts/audit_skills.py && uv run pytest tests/ -->