Skip to content

K-Dense-AI/scientific-agent-skills / skills/paper-lookup

paper-lookup

Flagged

Searches 18 scholarly APIs for papers, preprints, citations, open-access full text, repository records, and journal OA status, and returns results with reproducible provenance. Covers PubMed, PMC, Europe PMC, bioRxiv, medRxiv, arXiv, OpenAlex, Crossref, Semantic Scholar, CORE, Unpaywall, OpenCitations, PubTator3, Zenodo, Figshare, ROR, BioStudies, and DOAJ. Use when searching for papers, citations, DOI/PMID/arXiv lookups, abstracts, full text, open-access PDFs, preprints, citation graphs, author publications, biomedical entity annotations, deposited records (Zenodo, Figshare, BioStudies), institution ROR IDs, or any scholarly literature query. Triggers on mentions of any supported database or requests like "find papers on X", "look up this DOI", "who cites this paper", or "get me the PDF".

FileKindSize
SKILL.mdskill md25.5 KB
references/arxiv.mdmarkdown11.9 KB
references/biorxiv.mdmarkdown6.7 KB
references/biostudies.mdmarkdown4.0 KB
references/core.mdmarkdown4.6 KB
references/crossref.mdmarkdown5.5 KB
references/doaj.mdmarkdown4.2 KB
references/europepmc.mdmarkdown9.0 KB
references/figshare.mdmarkdown3.1 KB
references/medrxiv.mdmarkdown5.0 KB
references/openalex.mdmarkdown5.5 KB
references/opencitations.mdmarkdown4.6 KB
references/pmc.mdmarkdown7.1 KB
references/pubmed.mdmarkdown4.5 KB
references/pubtator.mdmarkdown4.5 KB
references/ror.mdmarkdown4.2 KB
references/semantic-scholar.mdmarkdown7.1 KB
references/unpaywall.mdmarkdown3.5 KB
references/zenodo.mdmarkdown3.8 KB
scripts/_common.pyscript8.2 KB
scripts/arxiv_atom.pyscript8.0 KB
scripts/jats_to_text.pyscript11.0 KB
scripts/openalex_abstract.pyscript6.3 KB
scripts/paginate.pyscript19.2 KB

references/pubtator.md

View on GitHub
# PubTator3

NCBI text-mined annotations on PubMed abstracts (and some PMC full text): genes,
diseases, chemicals, species, variants, and cell lines, plus typed relations.
Use it when the user wants *entities in a paper* or papers that mention a
normalized entity (`@CHEMICAL_remdesivir`), not when they want a citation list.

Europe PMC has a separate Annotations API. PubMed search
does not normalize "remdesivir" to a concept ID. PubTator3 APIs are **not** the
old PubTator / `CBBresearch` endpoints.

All figures below verified 2026-09-10.

## Base URL

```
https://www.ncbi.nlm.nih.gov/research/pubtator3-api
```

Docs: https://www.ncbi.nlm.nih.gov/research/pubtator3/api

## Authentication

None.

## Rate Limits

No more than **3 requests per second**. Serialize. For bulk annotation dumps use
the FTP site (`https://ftp.ncbi.nlm.nih.gov/pub/lu/PubTator3/`) rather than the
API.

## Key Endpoints

### 1. Resolve a mention to an entity ID

```
GET /entity/autocomplete/?query={text}&concept={type}&limit={n}
```

`concept` is optional (`chemical`, `disease`, `gene`, `species`, `variant`,
`cellline`).

```
GET /entity/autocomplete/?query=remdesivir&limit=3
```

Returns a JSON array. First hit (verified):

```json
{
  "_id": "@CHEMICAL_remdesivir",
  "biotype": "chemical",
  "db_id": "C000606551",
  "db": "ncbi_mesh",
  "name": "remdesivir"
}
```

Search with `_id` (`@CHEMICAL_remdesivir`), not the display name. Nearby
synonyms (`@CHEMICAL_GS_441524_triphosphate`) are different entities.

### 2. Search papers by text, entity, or relation

```
GET /search/?text={query}&page={n}
```

`text` may be free text, an `@TYPE_name` entity ID, a boolean combination, or a
relation:

```
@CHEMICAL_Doxorubicin AND @DISEASE_Neoplasms
relations:ANY|@CHEMICAL_Doxorubicin|@DISEASE_Neoplasms
relations:ANY|@CHEMICAL_Doxorubicin|DISEASE
```

```
GET /search/?text=@CHEMICAL_remdesivir
```

Verified: `count` 23295, `page_size` 10, `current` 1, `total_pages` 2330,
`results` length 10. First result `pmid` is an **integer** (`37711410`); `_id`
is a string. Page with `page` (1-based). There is no cursor.

Do not use this as a general PubMed replacement. Rank is entity-centric.

### 3. Export annotations for PMIDs

```
GET /publications/export/{format}?pmids={id,id}&full={true|false}
```

`format` is `pubtator`, `biocxml`, or `biocjson`. `full=true` (PMC full text)
works only for `biocxml` / `biocjson`.

```
GET /publications/export/biocjson?pmids=29355051
```

The JSON is **not** a bare BioC document. It is:

```json
{ "PubTator3": [ { "_id": "29355051|None", "id": "29355051", "passages": [...], "relations": [...] } ] }
```

Verified on PMID 29355051: one document, two passages, first passage has 5
annotations. An annotation looks like:

```json
{
  "infons": {
    "type": "Species",
    "database": "ncbi_taxonomy",
    "normalized_id": 112863,
    "biotype": "species"
  },
  "text": "Lycium barbarum",
  "locations": [{"offset": 14, "length": 15}]
}
```

Read `PubTator3[0].passages[].annotations`. A 200 with `"PubTator3": []` is
"no documents," not a transport success you can ignore.

### 4. Related entities

```
GET /relations?e1={entityId}&type={relation}&e2={entity_type}
```

Relation types include `treat`, `cause`, `interact`, `associate`,
`positive_correlate`, `negative_correlate`, `prevent`, `inhibit`, `stimulate`,
`drug_interact`, `cotreat`, `convert`, and `compare`. `e1` must be an autocomplete `_id`; `type` and `e2` are optional. Related entity types include gene, disease, chemical, and variant.

## Typical Workflow

1. Autocomplete the user's string → `@CHEMICAL_…` / `@DISEASE_…`.
2. Search with that ID (and a relation if they asked "what does X treat?").
3. Export `biocjson` for the PMIDs you will report, and list the annotations.
4. For the paper itself (abstract, OA PDF), go to PubMed / Europe PMC /
   Unpaywall. PubTator full-text annotations cover a subset; check returned passages before claiming full-text coverage.

## Failure Modes

| What you did | What happens | What to do |
|---|---|---|
| Parsed export JSON as BioC root | No `passages` at the top level | Descend into `PubTator3` |
| Searched `remdesivir` and treated hits as exact-chemical papers | Keyword search, not the concept | Autocomplete, then search `@CHEMICAL_remdesivir` |
| Called the old `pubtator-api` or `CBBresearch` URL | May still 200, but the contract changed | Use `pubtator3-api` |
| `full=true` with `format=pubtator` | No full text | Use `biocjson` or `biocxml` |
| Parallel fan-out | Easy to exceed 3 req/s | Serialize |

## Official sources reviewed 2026-09-30

- https://www.ncbi.nlm.nih.gov/research/pubtator3/api
deprecation-and-migrationaddyosmani/agent-skillsManages deprecation and migration. Use when removing old systems, APIs, or features. Use when migrating users from one implementation to another. Use when migrating a database schema in production, such as renaming or dropping a column without downtime (expand/contract). Use when deciding whether to maintain or sunset existing code.Stars: 119kUpdated Claude CodeCodexVerifiedInstall
performance-optimizationaddyosmani/agent-skillsOptimizes application performance across frontend, backend, queries, and databases. Use when performance requirements exist, when you suspect performance regressions, when Core Web Vitals or load times need improvement, when N+1 query patterns need fixing, or when profiling reveals bottlenecks.Stars: 119kUpdated Claude CodeCodexVerifiedInstall
agent-cost-reportthedotmack/claude-memBelievable agent cost report for any period, default the last 7 full days PT, not counting today. Measured tokens from Claude Code transcripts priced at OpenRouter list prices (ESTIMATED), measured provider spend when a sanctioned source exists, note-taker cost separate, Timing-style HTML/PDF plus report.json, line-items.csv, evidence.json.Stars: 118kUpdated Claude CodeCodexVerifiedInstall
agent-cost-reportthedotmack/claude-memBelievable agent cost report for any period, default the last 7 full days PT, not counting today. Measured tokens from Claude Code transcripts priced at OpenRouter list prices (ESTIMATED), measured provider spend when a sanctioned source exists, note-taker cost separate, Timing-style HTML/PDF plus report.json, line-items.csv, evidence.json.Stars: 118kUpdated Claude CodeCodexVerifiedInstall