Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Autonomous curator for Source Library - discover, evaluate, and import historical texts in alchemy, Hermetica, Kabbalah, Rosicrucianism, and early modern knowledge. Use when asked to curate books, find new sources, expand the collection, or build thematic batches.
.claude/skills/majiayu000-library-curator/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-01 | ✗→✓ | ▲ Improved | 7% | 0% |
| case-02 | ✗→✓ | ▲ Improved | 84% | 0% |
| case-03 | ✗→✓ | ▲ Improved | 25% | 0% |
| case-07 | ✗→✓ | ▲ Improved | 28% | 0% |
| case-09 | ✗→✓ | ▲ Improved | 135% | 0% |
Autonomous book acquisition agent for Source Library, focused on Western esoteric tradition and early modern knowledge.
Affiliation: Embassy of the Free Mind (Bibliotheca Philosophica Hermetica, Amsterdam) Mission: Build a comprehensive digital library of Western esoteric tradition and early modern knowledge
| Criterion | Weight | Notes | |-----------|--------|-------| | Thematic fit | 3x | Core esoteric tradition | | Edition quality | 2x | First editions, important printings | | Rarity | 2x | Not widely available digitally | | Historical authenticity | 2x | Original vs modern editions | | Completeness | 1x | Full text vs fragments | | Image quality | 1x | Readable scans | | Research value | 1x | Citations, scholarly interest |
bashcurl -X POST "https://sourcelibrary.org/api/import/ia" \ -H "Content-Type: application/json" \ -d '{ "ia_identifier": "bookid123", "title": "Book Title", "author": "Author Name", "year": 1617, "original_language": "Latin" }'
bashcurl -X POST "https://sourcelibrary.org/api/import/gallica" \ -H "Content-Type: application/json" \ -d '{ "ark": "bpt6k61073880", "title": "Book Title", "author": "Author Name", "year": 1617, "original_language": "Latin" }'
bashcurl -X POST "https://sourcelibrary.org/api/import/mdz" \ -H "Content-Type: application/json" \ -d '{ "bsb_id": "bsb00029099", "title": "Book Title", "author": "Author Name", "year": 1473, "original_language": "Latin" }'
Before importing, always check if the book is already in the collection:
bash# Search by title curl -s "https://sourcelibrary.org/api/search?q=TITLE" # Get all books curl -s "https://sourcelibrary.org/api/books" | jq '.[] | {id, title, author, year}' # Search by author curl -s "https://sourcelibrary.org/api/books" | jq '.[] | select(.author | contains("AUTHOR_NAME"))'
data/bph_catalog.csv (28,814 entries)data/ia_catalog.csv (9,000 entries)## [Title] ([Year])
**Author**: [Name]
**Language**: [Lang] | **Pages**: [N] | **Source**: [Archive ID]
**Theme**: [Primary collection]
**Score**: [N]/10
**Notes**: [1-2 sentences on significance]
**Status**: [acquired/skipped/pending]# Acquisition Batch [DATE] - [THEME]
## Summary
- Books acquired: N
- Total pages: N
- Languages: X, Y, Z
- Date range: YYYY-YYYY
## Thematic Rationale
[Why this batch, how it connects]
## Books
[Individual reports]
## Gaps Identified for Future Batches
[What to acquire next]FLAG:OCR - OCR quality problemsFLAG:ALIGN - Image/text misalignmentFLAG:META - Metadata errorsFLAG:INCOMPLETE - Missing pagesFLAG:DUPLICATE - Already in collection| Author | What We Need | Priority | |--------|--------------|----------| | Thomas Vaughan | Lumen de Lumine, Aula Lucis, Anima Magica Abscondita | HIGH | | Gichtel | Theosophia Practica | HIGH | | Jane Lead | English Philadelphian Society | MEDIUM | | Cudworth | True Intellectual System | MEDIUM |
| Author/Text | Have | Need | |-------------|------|------| | Boehme | 3 works | More German originals (Aurora, Signatura Rerum) | | Fludd | 3 works | Complete Utriusque Cosmi (5+ volumes) | | Dee | 1 work | True Relation, Monas hieroglyphica | | Paracelsus | Several | Individual treatises in German |
agentcurator.mdagentcurator.mdcuratorreports.md| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | fail→pass | 30,462 | 23,185 | -24% | 1 | 1 | 0% | 5,444 | 5,814 | +7% | 0 | 0 | — |
case-02 | fail→pass | 28,286 | 37,442 | +32% | 1 | 1 | 0% | 4,765 | 8,789 | +84% | 0 | 0 | — |
case-03 | fail→pass | 32,807 | 24,663 | -25% | 1 | 1 | 0% | 5,246 | 6,580 | +25% | 0 | 0 | — |
case-04 | pass→pass | 40,749 | 41,387 | +2% | 1 | 1 | 0% | 5,724 | 7,875 | +38% | 0 | 0 | — |
case-05 | pass→pass | 23,778 | 14,112 | -41% | 1 | 1 | 0% | 3,063 | 4,099 | +34% | 0 | 0 | — |
case-06 | pass→pass | 36,240 | 51,684 | +43% | 1 | 1 | 0% | 5,388 | 10,348 | +92% | 0 | 0 | — |
case-07 | fail→pass | 20,256 | 10,837 | -46% | 1 | 1 | 0% | 2,421 | 3,105 | +28% | 0 | 0 | — |
case-08 | fail→fail | 17,444 | 3,521 | -80% | 1 | 1 | 0% | 2,304 | 2,820 | +22% | 0 | 0 | — |
case-09 | fail→pass | 6,677 | 9,392 | +41% | 1 | 1 | 0% | 1,288 | 3,026 | +135% | 0 | 0 | — |
case-10 | fail→pass | 6,202 | 6,374 | +3% | 1 | 1 | 0% | 1,283 | 3,291 | +157% | 0 | 0 | — |
case-11 | fail→pass | 12,329 | 10,670 | -13% | 1 | 1 | 0% | 1,419 | 3,214 | +126% | 0 | 0 | — |
case-12 | fail→pass | 17,885 | 6,878 | -62% | 1 | 1 | 0% | 2,281 | 3,452 | +51% | 0 | 0 | — |
case-13 | pass→pass | 10,884 | 6,181 | -43% | 1 | 1 | 0% | 1,781 | 3,223 | +81% | 0 | 0 | — |
case-14 | pass→pass | 22,106 | 23,939 | +8% | 1 | 1 | 0% | 3,171 | 5,531 | +74% | 0 | 0 | — |
case-15 | fail→pass | 17,829 | 8,972 | -50% | 1 | 1 | 0% | 1,905 | 2,788 | +46% | 0 | 0 | — |
case-16 | pass→pass | 8,380 | 4,316 | -48% | 1 | 1 | 0% | 1,359 | 2,882 | +112% | 0 | 0 | — |
case-17 | fail→pass | 11,591 | 2,353 | -80% | 1 | 1 | 0% | 2,055 | 2,450 | +19% | 0 | 0 | — |
case-18 | fail→pass | 7,989 | 7,163 | -10% | 1 | 1 | 0% | 1,374 | 2,495 | +82% | 0 | 0 | — |
case-19 | pass→pass | 9,039 | 2,974 | -67% | 1 | 1 | 0% | 1,507 | 2,630 | +75% | 0 | 0 | — |
case-20 | fail→pass | 5,770 | 7,553 | +31% | 1 | 1 | 0% | 950 | 2,518 | +165% | 0 | 0 | — |
case-21 | pass→pass | 16,890 | 9,648 | -43% | 1 | 1 | 0% | 1,928 | 3,054 | +58% | 0 | 0 | — |
case-22 | fail→pass | 10,264 | 9,290 | -9% | 1 | 1 | 0% | 1,733 | 2,860 | +65% | 0 | 0 | — |
case-23 | fail→pass | 6,289 | 3,390 | -46% | 1 | 1 | 0% | 1,065 | 2,728 | +156% | 0 | 0 | — |
case-24 | pass→pass | 15,744 | 2,967 | -81% | 1 | 1 | 0% | 1,664 | 2,622 | +58% | 0 | 0 | — |
DecimalAI ran this skill against gemini-3.6-flash twice over the same eval suite — once with the skill loaded and once without — and compared the two runs case by case. 24 cases were attempted. The headline lift of +58 percentage points is the difference between those two pass rates over the 24 comparable cases.
Without the skill loaded, the model failed this case. With it loaded, the same prompt on the same model passed. This is one improved case from the latest verified run; every case, including any that regressed, is in the table above.
Other measured skills in the registry, with their headline benchmark lift.