agentsope/SkillAlchemy / skills/LEAP
LEAP
VerifiedLEAP builds skills through two pipelines: Branch A distills a skill from raw data, while Branch B combines multiple skills into one. It is called by the main SkillAlchemy workflow. Use when SkillAlchemy requires distillation or fusion.
| File | Kind | Size |
|---|---|---|
| SKILL.md | skill md | 30.5 KB |
| domains/cybersecurity/domain.md | markdown | 2.1 KB |
| domains/energy/domain.md | markdown | 2.1 KB |
| domains/finance/domain.md | markdown | 1.9 KB |
| domains/healthcare/domain.md | markdown | 2.1 KB |
| domains/manufacturing/domain.md | markdown | 2.1 KB |
| domains/mathematics/domain.md | markdown | 2.1 KB |
| domains/media-content-production/domain.md | markdown | 2.4 KB |
| domains/natural-science/domain.md | markdown | 2.0 KB |
| domains/office-white-collar/domain.md | markdown | 2.1 KB |
| domains/persona-os/domain.md | markdown | 5.3 KB |
| domains/robotics/domain.md | markdown | 1.9 KB |
| domains/software-engineering/domain.md | markdown | 2.0 KB |
| references/skill-grammar.md | markdown | 17.3 KB |
| scripts/build_component_index.py | script | 23.7 KB |
| scripts/build_corpus.py | script | 11.0 KB |
| scripts/download_subtitles.sh | script | 2.1 KB |
| scripts/score_skill.py | script | 6.8 KB |
| scripts/srt_to_transcript.py | script | 3.3 KB |
| skill.json | config | 132 B |
scripts/score_skill.py
View on GitHub#!/usr/bin/env python3
"""
Score a SKILL.md against the 13-point mechanical quality rubric.
Usage:
python3 score_skill.py --skill <path>/SKILL.md # single file
python3 score_skill.py --skill <path>/SKILL.md --json # JSON output
python3 score_skill.py --dir <path>/skills/ # batch directory
The 13-point mechanical rubric is derived from statistical analysis of 477 skills
on skills.sh. See references/skill-grammar.md for methodology.
The mechanical score is used only for exemplar filtering in A-Stage 5 and
skill retrieval in B-Step 1.
"""
from __future__ import annotations
import argparse
import json
import re
import sys
from pathlib import Path
def score_skill(skill_path: Path) -> dict:
"""Score a single SKILL.md and return breakdown."""
if not skill_path.exists():
return {"error": "file not found", "path": str(skill_path)}
text = skill_path.read_text(encoding="utf-8")
lines = text.split("\n")
body_lines = [l for l in lines if l.strip() and not l.strip().startswith("#")]
line_count = len(lines)
score = 0
strengths = []
weaknesses = []
# 1. frontmatter has name (1 pt)
has_name = bool(re.search(r"^name:\s*\S", text, re.MULTILINE))
if has_name:
score += 1
strengths.append("has_frontmatter_name")
else:
weaknesses.append("missing_frontmatter_name")
# 2. frontmatter has description (1 pt)
has_desc = bool(re.search(r"^description:\s*", text, re.MULTILINE))
if has_desc:
score += 1
strengths.append("has_description")
else:
weaknesses.append("missing_description")
# 3. description contains trigger words (2 pts, weighted)
desc_match = re.search(r"^description:\s*(.+?)(?=\n\S|\n$|\Z)", text, re.MULTILINE | re.DOTALL)
desc_text = desc_match.group(1) if desc_match else ""
trigger_keywords = [
"use when", "when user", "when the user", "trigger",
"activates when", "applies when"
]
has_trigger = any(kw.lower() in desc_text.lower() for kw in trigger_keywords)
if has_trigger:
score += 2
strengths.append("description_has_trigger")
else:
weaknesses.append("description_missing_trigger")
# 4. description is specific (>80 chars) (1 pt)
if len(desc_text.strip()) > 80:
score += 1
strengths.append("description_specific")
else:
weaknesses.append("description_too_vague")
# 5. >=3 major sections (1 pt)
h2_count = len(re.findall(r"^##\s+", text, re.MULTILINE))
h3_count = len(re.findall(r"^###\s+", text, re.MULTILINE))
if h2_count + h3_count >= 3:
score += 1
strengths.append("has_multiple_sections")
else:
weaknesses.append("too_few_sections")
# 6. has boundary declaration (2 pts, weighted)
boundary_patterns = [
r"boundary", r"boundaries", r"limitations?", r"constraints",
r"out of scope",
r"boundary rules", r"boundary rule"
]
has_boundary = any(re.search(p, text, re.IGNORECASE) for p in boundary_patterns)
if has_boundary:
score += 2
strengths.append("has_boundary_section")
else:
weaknesses.append("missing_boundary")
# 7. >=5 concrete steps (2 pts, weighted)
step_patterns = [
r"step\s*\d", r"Step\s*\d", r"phase\s*\d", r"Phase\s*\d",
r"^\d+\.\s",
]
step_count = 0
for pat in step_patterns:
step_count += len(re.findall(pat, text, re.MULTILINE))
if step_count >= 5:
score += 2
strengths.append("has_concrete_steps")
else:
weaknesses.append("insufficient_steps")
# 8. has examples section (1 pt)
has_examples = bool(re.search(
r"example|demo|conversation|scenario",
text, re.IGNORECASE
))
if has_examples:
score += 1
strengths.append("has_examples_section")
else:
weaknesses.append("missing_examples")
# 9. line count 100-400 (1 pt)
if 100 <= line_count <= 400:
score += 1
strengths.append("line_count_sweet_spot")
else:
weaknesses.append(f"line_count_{line_count}")
# 10. has references/related section (1 pt)
has_refs = bool(re.search(
r"reference|see also|related",
text, re.IGNORECASE
))
if has_refs:
score += 1
strengths.append("has_references_section")
else:
weaknesses.append("missing_references")
# 11. penalty: <30 lines (-2 pts)
if line_count < 30:
score -= 2
weaknesses.append("too_thin_penalty")
# Determine grade
if score >= 11:
grade = "elite"
elif score >= 9:
grade = "standard"
else:
grade = "draft"
# Determine skill mode
persona_signals = [
"role-playing rules", "how i speak", "things i would never say",
"signature phrase", "how i see the world", "decision heuristics", "persona"
]
lower_text = text.lower()
skill_mode = "persona" if any(signal in lower_text for signal in persona_signals) else "tool"
return {
"path": str(skill_path),
"score": min(score, 13),
"max": 13,
"grade": grade,
"skill_mode": skill_mode,
"line_count": line_count,
"strengths": strengths,
"weaknesses": weaknesses,
}
def main():
parser = argparse.ArgumentParser(description="Score SKILL.md files")
parser.add_argument("--skill", type=Path, help="Path to SKILL.md")
parser.add_argument("--dir", type=Path, help="Path to directory of SKILL.md files")
parser.add_argument("--json", action="store_true", help="Output as JSON")
args = parser.parse_args()
if args.skill:
result = score_skill(args.skill)
if args.json:
print(json.dumps(result, indent=2, ensure_ascii=False))
else:
print(f"{result['path']}: {result['score']}/13 ({result['grade']}) — {result['skill_mode']}")
if result.get("strengths"):
print(f" ✓ {', '.join(result['strengths'])}")
if result.get("weaknesses"):
print(f" ✗ {', '.join(result['weaknesses'])}")
elif args.dir:
results = []
for md in sorted(args.dir.glob("**/SKILL.md")):
results.append(score_skill(md))
results.sort(key=lambda r: r["score"], reverse=True)
if args.json:
print(json.dumps(results, indent=2, ensure_ascii=False))
else:
for r in results:
print(f"[{r['score']:2}/13 {r['grade']:8s}] {r['skill_mode']:6s} {r['path']}")
elite = sum(1 for r in results if r["grade"] == "elite")
standard = sum(1 for r in results if r["grade"] == "standard")
draft = sum(1 for r in results if r["grade"] == "draft")
print(f"\n{len(results)} skills: {elite} elite, {standard} standard, {draft} draft")
else:
parser.print_help()
if __name__ == "__main__":
main()
Related skills
| Skill | Stars | Updated | Works with | Status | Install |
|---|---|---|---|---|---|
| writing-skillsobra/superpowersUse when creating new skills, editing existing skills, or verifying skills work before deployment | Stars: 322k | Updated | Claude CodeCodex | Verified | Install |
| xlsxanthropics/skillsUse this skill any time a spreadsheet file is the primary input or output. This means any task where the user wants to: open, read, edit, or fix an existing .xlsx, .xlsm, .xltx, .csv, or .tsv file (e.g., adding columns, computing formulas, formatting, charting, cleaning messy data); create a new spreadsheet from scratch or from other data sources; or convert between tabular file formats. Trigger especially when the user references a spreadsheet file by name or path — even casually (like "the xlsx in my downloads") — and wants something done to it or produced from it. Also trigger for cleaning or restructuring messy tabular data files (malformed rows, misplaced headers, junk data) into proper spreadsheets. The deliverable must be a spreadsheet file. Do NOT trigger when the primary deliverable is a Word document, HTML report, standalone Python script, database pipeline, or Google Sheets API integration, even if tabular data is involved. | Stars: 197k | Updated | Claude CodeCodex | Verified | Install |
| shipping-and-launchaddyosmani/agent-skillsPrepares production launches. Use when preparing to deploy to production, or when asking what needs to be in place before shipping. Use when you need a pre-launch checklist, when setting up monitoring, when planning a staged rollout, or when you need a rollback strategy. | Stars: 119k | Updated | Claude CodeCodex | Verified | Install |
| childhoodthedotmack/claude-memBabysit a shipped release after merge, publication, or deployment. Verify published packages, production behavior, installed plugins, and late CI or user reports until the observation window is clean; use babysit for PR readiness before release. | Stars: 118k | Updated | Claude CodeCodex | Verified | Install |