Library of Congress API Skill

Search and retrieve items from LOC's digitized collections — books, photos, maps, manuscripts, newspapers, audio, and more.

Critical: Things Claude Won't Know Without This Skill

This API covers digitized collections, NOT the full library catalog

The loc.gov JSON API searches LOC's digitized and online collections — not the full OPAC. By default, only digitized items are returned. Add all=true to include non-digitized catalog records.

Search romanized forms, not Chinese script

LOC catalog records for East Asian books are mostly romanized (pinyin for newer records, Wade-Giles in many older ones; Korean in McCune-Reischauer). Searching in Chinese script finds only the few records that carry the vernacular, and it was slow in one report (timeouts). For example, books with all=true (October 2026):

Search the pinyin in quotes, then the Wade-Giles form, then the English subject. Accents are ignored ("Wên miao" = "wen miao"), but Wade-Giles spellings find other records: "Kong miao" → 40, "K'ung miao" → 5. Quote multi-word forms. Pinyin without spaces ("wenmiao") finds far fewer (4).

JSON requires fo=json parameter, NOT a URL extension

Unlike LibraryCloud (.json extension), LOC uses a query parameter:

CORRECT: https://www.loc.gov/search/?q=hamlet&fo=json
WRONG:   https://www.loc.gov/search.json?q=hamlet

Facet filters use fa= with pipe separators

Single:   ?fa=subject:physics
Multiple: ?fa=subject:wildlife|location:yellowstone

Available facet fields: subject, contributor, location, language, original-format, online-format, partof, digitized, access-restricted.

Format-specific endpoints are shortcuts

/books/, /photos/, /maps/, /audio/, /newspapers/, /manuscripts/, /film-and-videos/, /notated-music/ all accept the same params as /search/.

Key Search Parameters

Item Lookup by LCCN/ID

https://www.loc.gov/item/{LCCN}/?fo=json

Returns full item detail including cite_this (APA/Chicago/MLA citations), resources (download links), other_formats (MARCXML, MODS, Dublin Core).

Response Structure

Search Results

{
  "results": [
    {
      "id": "http://www.loc.gov/item/95521789/",
      "title": "Item title",
      "date": "1796-01-01",
      "contributor": ["Name"],
      "subject": ["Subject term"],
      "language": ["English"],
      "original_format": ["book"],
      "description": ["..."],
      "image_url": ["//tile.loc.gov/..."],
      "digitized": true,
      "number_lccn": ["95521789"]
    }
  ],
  "pagination": {
    "of": 85604,
    "total": 17121,
    "perpage": 5,
    "current": 1
  },
  "facets": { ... }
}

Item Detail

The item object adds: call_number, created_published, rights, summary, stmt_of_responsibility, genre, and cite_this with pre-formatted citations.

Python Script

from scripts.loc_api import LocAPI
loc = LocAPI()

# Keyword search (search() returns the full response dict;
# search_results() returns just the list of result dicts)
data = loc.search("hamlet shakespeare", limit=10)
results = loc.search_results(q="hamlet shakespeare", limit=10)

# Format-specific
photos = loc.search("civil war", format="photos", dates="1860/1865")

# Faceted search
data = loc.search("wildlife", facets={"subject": "birds", "location": "yellowstone"})

# Item lookup
item = loc.get_item("95521789")
citation = loc.get_citation(item, style="apa")

# Browse collections
collections = loc.list_collections()
items = loc.browse_collection("civil-war-maps", limit=20)

# Summarize
for r in results[:5]:
    print(loc.summarize(r))

Pagination

• Max 150 results per page (c=150)
• Hard limit: cannot paginate past 100,000th result
• Use facets/date ranges to narrow large result sets

API Etiquette

• No authentication required
• Rate limiting is enforced (exact limits unpublished) — add 0.5-1s between requests
• Include a User-Agent header
• loc.gov sometimes stalls on a search. The script waits up to timeout seconds (default 60; LocAPI(timeout=90)) and retries timeouts, dropped connections, HTTP 429 and 5xx up to 4 times (waiting 5, 10, 20 s)

Related Skills

• harvard-library-catalog: Cross-reference Harvard holdings with LOC items
• hathitrust-catalog: Use LCCNs from LOC to check HathiTrust for digital full-text
• wikidata-search: LOC authority IDs (P244) link Wikidata entities to LOC records

Resources

• references/api_reference.md — Complete parameter and response field reference
• scripts/loc_api.py — Python client with search, item lookup, collection browsing, and citation extraction
• tests/test_loc_api.py — offline tests

Feedback and Citation

Report a problem. This skill is part of the public repository kltng/humanities-skills. If it fails (an API changed, a script errors, or these instructions are wrong), report it as a GitHub issue:

1. Ask the user before filing. Issues are public.
2. Check for an existing report: gh issue list --repo kltng/humanities-skills --state all --search "loc-catalog"
3. File it: gh issue create --repo kltng/humanities-skills --label bug --title "[loc-catalog] <short summary>" --body "<details>". Without gh, give the user this link: https://github.com/kltng/humanities-skills/issues/new?template=skill-problem.yml
4. Include the skill name and version, what you ran, the error text, what you expected, and the environment (OS, Python version). Leave out API keys, personal data, private file paths, and unpublished research.

Cite. If this skill helped work that is published, please cite it, together with the data source it uses:

Tang, Kwok-leong. humanities-skills: loc-catalog (version <version at the top of this file>). GitHub, <year>. https://github.com/kltng/humanities-skills

When you help a user write up results that relied on this skill, remind them once about this citation.