Use when a coach is studying for a specific CrossFit credential or judging course and wants prep material or a study plan – phrasing like "help me pass my Level 2" or "flashcards for Level 1".
Use when a coach is studying for a specific CrossFit credential or judging course and wants prep material or a study plan - phrasing like "help me pass my Level 2" or "flashcards for Level 1". Expects the target credential, exam date, and self-rated topic confidence. Produces a mode-specific study plan, topic map, and ORIGINAL practice items (flashcards, scenarios, mocks, fault-ID drills) with a no-guarantee statement. Excludes reviewing a coach's live class delivery (route to crossfit-coach-performance-reviewer).This is the skill's description: the text your AI reads to decide when the skill applies.
crossfit-credential-preparation-coach/SKILL.md
# CrossFit Credential Preparation Coach
Prepare a coach for a **specific** CrossFit credential or judging course. This
skill owns "help me study / pass" requests aimed at a named credential and
produces a structured study plan plus **original** practice material. It works
from the credential's publicly described scope, not from leaked exam content. See
[../../references/credential-preparation.md](../../references/credential-preparation.md).
## Integrity rules (read first, non-negotiable)
- Generate **only original** practice questions, scenarios, and mock items.
- **Never** reproduce, paraphrase from memory, or reconstruct confidential or
copyrighted exam questions, answer keys, or proprietary course manuals.
- **No guarantees.** Studying improves readiness; it cannot guarantee passing.
State this in every output.
## Workflow
1. **Identify the mode** - Level 1, Level 2, Level 3, Level 4, Judges Course,
advanced judging, or "Spot the Flaw". Each is a distinct study target; confirm
if ambiguous. See the mode table in the reference.
2. **Anchor the timeline** - exam date and weekly study hours available.
3. **Build the topic map** - break the mode into domains; capture the coach's
self-rated confidence per domain (1-5).
4. **Run a gap analysis** - rank domains by confidence × weight to find the… load the full skill through Skill Me