Skip to content

api-reference.md

Latest commit

 

History

History
907 lines (712 loc) · 31.3 KB

api-reference.md

File metadata and controls

907 lines (712 loc) · 31.3 KB

Local API Reference

Language versions: 简体中文

Related documents: Overview · Quick Start · Plugin Development Guide

Basics

  • Protocol: HTTP
  • Default address: http://127.0.0.1:29467
  • API prefix: /api/v1
  • Data format: application/json; charset=utf-8
  • Supported API methods: GET, POST, PUT, OPTIONS
  • Write access: optional, disabled by default while Read-Only Mode is on, and only available after the user turns Read-Only Mode off in Settings → Local API. PUT /api/v1/papers/{uuid}/pdf additionally requires pdf-upload in /status.capabilities
  • Plugin static hosting lives under /plugins/{name}/... and is not under /api/v1

General Conventions

Successful responses

  • Normal business endpoints return 200 OK
  • Successful paper creation returns 201 Created
  • Preflight requests return 204 No Content

Error responses

Most error responses include at least:

{
  "error": "Error message"
}

Some error payloads include extra fields:

{
  "error": "Duplicate paper found",
  "existingPaperID": "550E8400-E29B-41D4-A716-446655440000"
}
{
  "error": "Library limit reached",
  "currentCount": 50,
  "limit": 50
}

Common status codes

Status Meaning Typical reason
200 Success The request completed normally
201 Created POST /api/v1/papers created a paper, or PUT /api/v1/papers/{uuid}/pdf uploaded bytes
204 No content OPTIONS preflight
400 Bad request Invalid request format, malformed JSON, empty title, invalid paperType, invalid strategy, missing metadata lookup query, or invalid PDF request shape
403 Forbidden Non-local request, origin not allowed, or public write access is still read-only
404 Not found Unknown endpoint, missing paper, failed metadata lookup, or missing plugin asset
405 Method not allowed Unsupported method such as PATCH or DELETE, or a method/path combination the API does not implement
409 Conflict Duplicate paper was detected during create, or a PDF is already attached and upload was not forced
422 Unprocessable Content Free-tier paper limit was reached
413 Payload too large Request body exceeded the configured size limit
503 Service unavailable PDF upload was requested but no usable upload destination is configured
500 Internal server error Lattice failed to process the request

Capability Detection

Use GET /api/v1/status before relying on optional behavior.

In particular:

  • treat create-paper in capabilities as the source of truth for whether POST /api/v1/papers is usable
  • treat create-paper-v2 as the signal that duplicate strategy and rich create/replace responses are supported
  • treat pdf-upload as the source of truth for whether PUT /api/v1/papers/{uuid}/pdf is usable
  • do not assume write access is enabled on every machine

Field-by-Method Matrix

This section is the quickest way to answer two questions:

  • which fields exist at all
  • which HTTP methods expose or accept each field

Status and capability fields

Field GET /status response Notes
ok ok Health flag
apiVersion apiVersion Local API version string
appVersion appVersion Lattice app version string
capabilities[] capabilities[] Capability list for the current instance
readOnly readOnly Whether the public /api/v1 write surface is currently read-only
baseURL baseURL Current localhost base address
browserExtensionEnabled browserExtensionEnabled Mirrors the Browser Extension toggle in Settings

Search fields

Field GET /search query GET /search response Notes
q q — Empty query returns recent papers
limit limit — Clamped to 1...50; unparsable values fall back to 10
papers[] — papers[] Top-level result array
id — papers[].id Search hit paper UUID
title — papers[].title Search hit display title
authorsDisplay — papers[].authorsDisplay Search-only author summary
subtitle — papers[].subtitle Search-only secondary summary
year — papers[].year Search hit year
citekey — papers[].citekey Search hit citation key
paperType — papers[].paperType Search hit paper type

Metadata lookup fields

Field GET /metadata query GET /metadata response Notes
doi doi doi At least one of doi, arxivId, isbn, or title is required
arxivId arxivId — Lookup input only
isbn isbn isbn Can be used as both lookup input and output
title title title Can be used as both lookup input and output
authors — authors Semicolon-separated author string
year — year Resolved publication year
journal — journal Resolved source / venue
abstract — abstract Resolved abstract text
volume — volume Resolved volume
issue — issue Resolved issue
pages — pages Resolved pages
paperType — paperType Resolved paper type when available

Collection directory fields

Field GET /collections response Notes
id [].id Collection UUID
name [].name Leaf collection name
path [].path Full slash-delimited collection path
depth [].depth Nesting depth for UI indentation

Tag directory fields

Field GET /tags response Notes
id [].id Tag UUID
name [].name Tag display name
colorHex [].colorHex Optional stored color

Core paper fields

Field GET /papers/{uuid} response GET /metadata response POST /papers request POST /papers response Notes
id id — — id Stable paper UUID; can be used to synthesize lattice://paper/{id}
citekey citekey — — — Read-only citation key
title title title title title Present in read and write flows
authors authors[] authors authors authors Detail uses string[]; metadata lookup and write request/response use a semicolon-separated string
year year year year year Shared core field
journal journal journal journal — Accepted on create, returned on detail and metadata lookup, not echoed by create response
abstract — abstract abstract — Stored on create and returned by metadata lookup, not currently exposed by paper-detail payloads
doi doi doi doi doi Shared core identifier
volume volume volume volume — Accepted on create, returned on detail and metadata lookup
issue issue issue issue — Accepted on create, returned on detail and metadata lookup
pages pages pages pages — Accepted on create, returned on detail and metadata lookup
isbn isbn isbn isbn — Accepted on create, returned on detail and metadata lookup
paperType paperType paperType paperType — Accepted on create, returned on detail, metadata lookup, and search
pdfPath — — pdfPath — Input-only field for path-based PDF attachment attempts during create
collections — — collections[] — Input-only case-insensitive collection-path list
tags — — tags[] — Input-only tag-name list
enrich — — enrich — Input-only background enrichment flag
strategy — — strategy — Duplicate strategy for create: "skip" or "replace"
pdfAttached — — — pdfAttached Whether the paper has an attached PDF after the operation
enrichmentStatus — — — enrichmentStatus Output-only write result field: pending or none
warnings — — — warnings[] Output-only non-fatal write warnings
alreadyExists — — — alreadyExists false for new creates, true when strategy: "replace" updated an existing paper
conflict — — — conflict Conflict summary object returned on successful replace responses
cslItem cslItem — — — Read-only CSL payload object

PDF upload fields

Field PUT /papers/{uuid}/pdf request PUT /papers/{uuid}/pdf response Notes
uuid path segment — Existing paper UUID
Content-Type application/pdf — Required
request body raw PDF bytes — Must start with %PDF-
force force=true query — Optional; bypasses the existing-PDF conflict check
reason — reason Success payload currently returns { "reason": "uploaded" }

CSL subfields

Field GET /papers/{uuid} response Notes
cslItem.id cslItem.id Same UUID as top-level id
cslItem.type cslItem.type CSL item type
cslItem.title cslItem.title CSL title
cslItem.author[] cslItem.author[] CSL author array
cslItem.author[].family cslItem.author[].family Present when the name can be split
cslItem.author[].given cslItem.author[].given Present when the name can be split
cslItem.author[].literal cslItem.author[].literal Used for mononyms, institutional names, or unsplittable names
cslItem.container-title cslItem.container-title Used for non-book container titles
cslItem.publisher cslItem.publisher Used for book-like items
cslItem.issued.date-parts cslItem.issued.date-parts Currently [[YYYY]] when a year exists
cslItem.DOI cslItem.DOI CSL DOI
cslItem.volume cslItem.volume CSL volume
cslItem.issue cslItem.issue CSL issue
cslItem.page cslItem.page CSL pages
cslItem.ISBN cslItem.ISBN CSL ISBN

Fields not currently exposed by read endpoints

These fields are not available in GET /search or GET /papers/{uuid} today:

  • abstract
  • collections
  • tags
  • notes
  • pdfPath
  • pdfURL
  • latticeURL

GET /status

Checks whether the service is online and reports which capabilities the current instance supports.

Request

GET /api/v1/status

Response fields

Field Type Description
ok boolean Health flag. Normally true when the service is available
apiVersion string Current Local API version
appVersion string Current Lattice app version
capabilities string[] Supported capability list
readOnly boolean Whether the public /api/v1 write surface is currently read-only
baseURL string Current localhost base URL
browserExtensionEnabled boolean Mirrors the Browser Extension toggle in Settings

Current capability values

Capability Description
search Paper search is available
paper-detail Per-paper detail fetch by paper ID is available
csl-item Detail payloads include CSL-JSON usable by citation engines
create-paper-v2 The create route supports duplicate strategy and rich create/replace responses
create-paper POST /api/v1/papers is enabled for this Lattice instance
pdf-upload PUT /api/v1/papers/{uuid}/pdf is currently available
plugin-hosting Local plugin static hosting is available

Example

{
  "ok": true,
  "apiVersion": "1",
  "appVersion": "1.2.3 (456)",
  "readOnly": false,
  "baseURL": "http://127.0.0.1:29467",
  "browserExtensionEnabled": true,
  "capabilities": [
    "search",
    "paper-detail",
    "csl-item",
    "plugin-hosting",
    "create-paper",
    "create-paper-v2",
    "pdf-upload"
  ]
}

GET /search

Searches the Lattice library and returns lightweight result cards suitable for suggestion UIs, pickers, and search result lists.

Request

GET /api/v1/search?q=<query>&limit=<n>

Query parameters

Parameter Required Type Default Description
q No string empty string Search query. If empty, the endpoint returns recently added papers
limit No integer 10 Requested result count. Valid integers are clamped to 1...50; unparsable values fall back to 10

Matching scope

From an external integration point of view, search matches the following kinds of fields:

  • title
  • author
  • journal / venue / source
  • citekey
  • year

Response shape

{
  "papers": [
    {
      "id": "550E8400-E29B-41D4-A716-446655440000",
      "title": "Attention Is All You Need",
      "authorsDisplay": "Ashish Vaswani, Noam Shazeer, Niki Parmar, ...",
      "subtitle": "Ashish Vaswani, Noam Shazeer, Niki Parmar, ... • 2017 • NeurIPS",
      "year": 2017,
      "citekey": "vaswani2017attention",
      "paperType": "inproceedings"
    }
  ]
}

papers[] field reference

Field Type Description
id string Paper UUID
title string Display title. Empty titles are returned as Untitled
authorsDisplay string Author text formatted for direct UI display
subtitle string Secondary summary text, typically combining author, year, and source
year integer | null Publication year
citekey string Preferred citation key. If no persisted key exists, a usable generated key is returned
paperType string Paper type

Possible paperType values

  • article
  • book
  • inproceedings
  • thesis
  • report
  • misc

Example

curl "http://127.0.0.1:29467/api/v1/search?q=transformer&limit=5"

GET /papers/{uuid}

Fetches detailed citation metadata for a single paper.

Request

GET /api/v1/papers/{uuid}

Path parameters

Parameter Type Description
uuid string Paper UUID. If the path segment is not a valid UUID, the request falls through to 404 Not Found

Response shape

{
  "id": "550E8400-E29B-41D4-A716-446655440000",
  "citekey": "vaswani2017attention",
  "title": "Attention Is All You Need",
  "authors": [
    "Ashish Vaswani",
    "Noam Shazeer"
  ],
  "year": 2017,
  "journal": "NeurIPS",
  "doi": "10.5555/example-doi",
  "volume": "30",
  "issue": null,
  "pages": "5998-6008",
  "isbn": null,
  "paperType": "inproceedings",
  "cslItem": {
    "id": "550E8400-E29B-41D4-A716-446655440000",
    "type": "paper-conference",
    "title": "Attention Is All You Need",
    "author": [
      {
        "family": "Vaswani",
        "given": "Ashish"
      },
      {
        "family": "Shazeer",
        "given": "Noam"
      }
    ],
    "container-title": "NeurIPS",
    "publisher": null,
    "issued": {
      "date-parts": [[2017]]
    },
    "DOI": "10.5555/example-doi",
    "volume": "30",
    "issue": null,
    "page": "5998-6008",
    "ISBN": null
  }
}

Top-level field reference

Field Type Description
id string Paper UUID
citekey string Preferred citation key
title string Title
authors string[] Array of author strings
year integer | null Year
journal string | null Journal, conference, publisher, or other source text
doi string | null DOI
volume string | null Volume
issue string | null Issue
pages string | null Page range
isbn string | null ISBN
paperType string Paper type
cslItem object Object ready to be passed to a CSL citation processor

cslItem field reference

Field Type Description
id string Same as the paper ID
type string CSL item type
title string Title
author object[] Author array with family / given / literal fields
container-title string | null Journal, proceedings, or other container title
publisher string | null Publisher. Common for book-like items
issued object | null Date object in the form {"date-parts": [[YYYY]]}
DOI string | null DOI
volume string | null Volume
issue string | null Issue
page string | null Pages
ISBN string | null ISBN

cslItem.author[] field reference

Field Type Description
family string | null Family name when Lattice can split the author name
given string | null Given name when Lattice can split the author name
literal string | null Literal name used for one-part, institutional, or otherwise unsplittable names

cslItem.issued field reference

Field Type Description
date-parts integer[][] Currently [[YYYY]] when a year exists

paperType to cslItem.type mapping

paperType cslItem.type
article article-journal
book book
inproceedings paper-conference
thesis thesis
report report
misc article

Fields intentionally not returned

GET /api/v1/papers/{uuid} is citation-oriented. It currently does not expose:

  • abstract
  • collections
  • tags
  • notes
  • pdfPath
  • pdfURL
  • latticeURL

Deep-link note

If your integration only needs to open the paper in Lattice, you can synthesize:

lattice://paper/{id}

The current Local API does not return a dedicated latticeURL field, and it does not expose the local PDF file path.

Typical uses

  • Fetch a full paper snapshot before inserting a citation
  • Build input for citeproc / CSL engines
  • Show a detailed citation card in an external tool
  • Maintain a local metadata cache inside a plugin

GET /collections

Returns assignable collections for write-oriented clients.

Request

GET /api/v1/collections

Response shape

[
  {
    "id": "550E8400-E29B-41D4-A716-446655440000",
    "name": "Transformers",
    "path": "Deep Learning/Transformers",
    "depth": 1
  }
]

Behavioral notes

  • the response is a top-level JSON array, not an object wrapper
  • collections are sorted by path using localized case-insensitive comparison
  • use path as the write-side identifier for POST /api/v1/papers

GET /tags

Returns assignable tags for write-oriented clients.

Request

GET /api/v1/tags

Response shape

[
  {
    "id": "550E8400-E29B-41D4-A716-446655440000",
    "name": "foundational",
    "colorHex": "#FF9500"
  }
]

Behavioral notes

  • the response is a top-level JSON array, not an object wrapper
  • tags are sorted by stored sortOrder first and by name second
  • use name as the write-side identifier for POST /api/v1/papers

GET /metadata

Looks up citation metadata from a lightweight identifier.

Request

GET /api/v1/metadata?doi=<doi>
GET /api/v1/metadata?arxivId=<id>
GET /api/v1/metadata?isbn=<isbn>
GET /api/v1/metadata?title=<title>

Query parameters

Parameter Required Type Description
doi Conditionally string DOI lookup input
arxivId Conditionally string arXiv ID lookup input
isbn Conditionally string ISBN lookup input
title Conditionally string Title lookup input

At least one of these parameters must be present and non-empty.

Response shape

{
  "title": "Attention Is All You Need",
  "authors": "Ashish Vaswani; Noam Shazeer; Niki Parmar",
  "year": 2017,
  "journal": "NeurIPS",
  "abstract": "The dominant sequence transduction models are based on...",
  "doi": "10.48550/arXiv.1706.03762",
  "volume": "30",
  "issue": null,
  "pages": "5998-6008",
  "isbn": null,
  "paperType": "inproceedings"
}

Error responses

Status Typical payload Meaning
400 {"error":"Provide at least one of doi, arxivId, isbn, or title."} No lookup key was provided
404 {"error":"Not found"} No metadata was resolved
500 {"error":"Internal server error"} Encoding or server-side processing failed unexpectedly

Typical uses

  • Prefill a create form from a DOI, arXiv ID, ISBN, or title
  • Enrich a host application's lightweight search result before the user commits a write
  • Resolve a better title / author list before deciding whether a duplicate should be skipped or replaced

POST /papers

Creates one paper in the library.

Availability

  • The endpoint path is POST /api/v1/papers
  • It is intended for localhost callers only
  • It only works when /status.capabilities contains create-paper
  • If write access is disabled, the endpoint returns 403

Request body

{
  "title": "Attention Is All You Need",
  "authors": "Vaswani, Ashish; Shazeer, Noam; Parmar, Niki",
  "year": 2017,
  "journal": "NeurIPS",
  "abstract": "The dominant sequence transduction models are based on...",
  "doi": "10.48550/arXiv.1706.03762",
  "volume": "30",
  "issue": null,
  "pages": "5998-6008",
  "isbn": "9780306406157",
  "paperType": "inproceedings",
  "pdfPath": "/Users/me/Papers/attention.pdf",
  "collections": ["Deep Learning/Transformers"],
  "tags": ["foundational", "NLP"],
  "strategy": "skip",
  "enrich": true
}

Request body field reference

Field Required Type Default Description
title Yes string none Paper title. Leading/trailing whitespace is trimmed. Empty or whitespace-only titles are rejected
authors No string | null null Author list as a single semicolon-separated string. Example: "Einstein, Albert; Bohr, Niels"
year No integer | null null Publication year
journal No string | null null Journal, venue, publisher, or other source text
abstract No string | null null Abstract text. Stored on the paper, but not returned by the current read endpoints
doi No string | null null DOI. Lattice cleans DOI URLs to the canonical DOI form before saving
volume No string | null null Volume
issue No string | null null Issue
pages No string | null null Page range
isbn No string | null null ISBN. Cleaned and validated before saving; invalid values are dropped
paperType No string | null "misc" Paper type. Matching is case-insensitive. Invalid non-empty values return 400
pdfPath No string | null null Local filesystem path to a PDF to attach. The file must exist, be a PDF, and live inside a Trusted Folder or one of its subfolders. This is an input-only field; it is not returned by GET /papers/{uuid}
collections No string[] | null [] Collection paths to assign, using existing collection paths such as "AI/Transformers". Matching is case-insensitive
tags No string[] | null [] Tag names to attach. Existing tags are reused; missing tags are created automatically
strategy No string | null "skip" Duplicate handling strategy. "skip" returns 409 on duplicate; "replace" updates the existing paper and returns a success response
enrich No boolean | null false When true, Lattice queues background enrichment if the created paper has a DOI or an attached PDF

Request limits

  • title: at most 1024 UTF-8 bytes
  • abstract: at most 65536 UTF-8 bytes
  • authors: at most 16384 UTF-8 bytes
  • collections: at most 100 items
  • tags: at most 100 items
  • each collections[i] and tags[i]: at most 256 UTF-8 bytes

Possible paperType values

  • article
  • book
  • inproceedings
  • thesis
  • report
  • misc

Successful response

Status: 201 Created

{
  "id": "550E8400-E29B-41D4-A716-446655440000",
  "title": "Attention Is All You Need",
  "authors": "Vaswani, Ashish; Shazeer, Noam; Parmar, Niki",
  "year": 2017,
  "doi": "10.48550/arXiv.1706.03762",
  "pdfAttached": true,
  "enrichmentStatus": "pending",
  "warnings": ["Collection not found: Deep Learning/Missing"],
  "alreadyExists": false,
  "conflict": null
}

When strategy: "replace" updates an existing paper, the status is 200 OK and alreadyExists becomes true.

Successful response field reference

Field Type Description
id string Newly created paper UUID
title string Saved display title
authors string | null Saved authors joined with "; "
year integer | null Saved year
doi string | null Cleaned DOI
pdfAttached boolean Whether the paper has an attached PDF after the operation
enrichmentStatus string "pending" when enrichment was queued, otherwise "none"
warnings string[] Non-fatal issues such as unknown collections or PDF attachment failures
alreadyExists boolean false for new creates, true when strategy: "replace" updated an existing paper
conflict object | null Conflict summary describing the matched existing paper when replace succeeded

Error responses

Status Typical payload Meaning
400 {"error":"Malformed JSON"} Invalid JSON request body
400 {"error":"Title is required and must not be empty"} Missing or blank title
400 {"error":"Invalid paperType"} Unsupported paperType value
400 {"error":"Invalid strategy"} Unsupported strategy value
403 {"error":"API is in read-only mode"} Read-Only Mode is still on
404 {"error":"Not found"} Path other than /api/v1/papers
409 {"error":"Duplicate paper found","existingPaperID":"uuid","existingTitle":"...","conflictField":"doi","latticeURL":"lattice://paper/..."} Duplicate DOI or normalized title match was found while strategy resolved to "skip"
422 {"error":"Library limit reached","currentCount":50,"limit":50} Free-tier paper limit was reached
500 {"error":"Failed to save"} Save failed unexpectedly

Behavioral notes

  • Duplicate detection checks DOI first and normalized title second
  • strategy defaults to "skip"
  • "replace" overwrites fields the incoming request actually supplies, but preserves existing values for omitted optional fields
  • collections are matched against existing collection paths case-insensitively; missing collections become warnings and are not auto-created
  • tags are matched by normalized name; missing tags are created automatically
  • pdfPath must point to an existing PDF inside a Trusted Folder or one of its subfolders; failures become warnings and do not abort paper creation
  • If pdfPath is outside every Trusted Folder, the warning is: PDF path is not inside a Trusted Folder. Add the enclosing folder in Settings > Local API > Trusted Folders, or attach the PDF manually in Lattice.
  • Lattice does not copy the PDF file into app storage; it stores a security-scoped bookmark instead
  • enrich: true only queues background enrichment when the created paper has a DOI or an attached PDF
  • After creation, GET /api/v1/papers/{uuid} still returns the citation-oriented read payload, not a full echo of all write-time fields

Minimal example

curl -X POST http://127.0.0.1:29467/api/v1/papers \
  -H "Content-Type: application/json" \
  -d '{"title":"A Brief History of Time"}'

PUT /papers/{uuid}/pdf

Uploads raw PDF bytes to an existing paper.

Availability

  • The endpoint path is PUT /api/v1/papers/{uuid}/pdf
  • It is intended for localhost callers only
  • Treat pdf-upload in /status.capabilities as the source of truth for whether the endpoint is usable
  • If write access is disabled, the endpoint returns 403

Request

PUT /api/v1/papers/{uuid}/pdf
Content-Type: application/pdf

Request body: raw PDF bytes.

Optional query parameter:

Parameter Required Type Default Description
force No boolean false When true, bypasses the existing-PDF conflict check

Successful response

Status: 201 Created

{
  "reason": "uploaded"
}

Error responses

Status Typical payload Meaning
400 {"error":"Content-Type must be application/pdf"} Wrong or missing content type
400 {"error":"Invalid PDF payload"} Body did not look like a PDF
403 {"error":"API is in read-only mode"} Read-Only Mode is still on
404 {"error":"Not found"} Unknown paper or unknown path
409 {"error":"PDF already attached"} The paper already has a PDF and force=true was not supplied
413 {"error":"Payload too large"} The PDF body exceeded the upload limit
503 {"reason":"not-configured"} No usable PDF upload destination is currently configured
500 {"error":"..."} Upload failed unexpectedly

Behavioral notes

  • the request body must be raw bytes and begin with %PDF-
  • the current upload limit for this route is 200000000 bytes
  • a successful upload updates the paper's stored PDF bookmark and may trigger a background refresh

GET /plugins/{name}/...

Serves static files from Lattice's local plugin directory.

Characteristics

  • Endpoint prefix: /plugins/
  • Purpose: host local front-end assets for plugins and add-ins
  • Not part of /api/v1
  • Static files are served from Lattice's sandboxed Application Support plugin directory
  • Path traversal attempts are rejected

Common content types

  • .html
  • .js
  • .css
  • .json
  • .png
  • .svg
  • .xml
  • .csl

OPTIONS

The Local API also accepts OPTIONS to support browser preflight requests.

Characteristics

  • Returns 204 No Content
  • Returns CORS headers for allowed origins
  • Does not return business payload data

If your page is hosted by Lattice under /plugins/{name}/..., you usually do not need to handle this explicitly.

CORS and origin restrictions

By default, the Local API is intended for local browser origins:

  • http://localhost:*
  • https://localhost:*
  • http://127.0.0.1:*
  • https://127.0.0.1:*

This is why hosting your plugin page under /plugins/{name}/... is the simplest and most reliable option.

Advanced note:

  • the server allows GET, POST, PUT, OPTIONS
  • the CORS allowlist can be extended with the citationBridgeAllowedOrigins user-defaults override when needed for advanced integrations

Request size limit

For most routes, the total HTTP request size is limited to 262144 bytes (256 KiB), including headers and body.

PUT /api/v1/papers/{uuid}/pdf has a larger PDF body limit of 200000000 bytes.

Error examples

Invalid paper path segment

curl http://127.0.0.1:29467/api/v1/papers/not-a-uuid
{
  "error": "Not found"
}

Missing metadata lookup key

curl http://127.0.0.1:29467/api/v1/metadata
{
  "error": "Provide at least one of doi, arxivId, isbn, or title."
}

Unknown endpoint

{
  "error": "Not found"
}

Related documents

  • If you have not yet verified that the service is reachable, start with Quick Start
  • If you are building a local plugin or host extension, pair this with Plugin Development Guide