bibcode
Enable an AI agent to recognise a bibcode, establish which bibliographic record it identifies, assess the reliability of that identification, and choose justified lookup, linking or correction actions.
Research draft, second pass
A second pass drafted this model: the structure a model of this thing needs, and what is known about it in the world. The line under this one says how the second half was obtained - researched against sources, or recalled without web access, in which case nothing here was read anywhere and every claim is a lead to verify. Unreviewed either way.
Researched by: Codex
Purpose and description
Enable an AI agent to recognise a bibcode, establish which bibliographic record it identifies, assess the reliability of that identification, and choose justified lookup, linking or correction actions.
It can be Extract a candidate bibcode while retaining its original citation or URL context.; Parse its positions using a documented convention and flag unsupported interpretations.; Query a bibliographic service for the exact identifier and record the outcome.; Compare returned metadata with the intended citation before attaching the identifier.; Follow evidenced identifier relationships while preserving the supplied code.; Propose a correction or replacement for review when lookup and citation evidence justify it..
Distinguishing features
A conventional bibcode has 19 characters arranged as YYYYJJJJJVVVVMPPPPA, with periods used for padding; appearance alone does not establish assignment. See [ADS bibcode documentation](https://ui.adsabs.harvard.edu/help/actions/bibcode).
Its conventional positions encode publication year, publication abbreviation, volume or publication type, qualifier, page-related information and first-author surname initial. See [ADS format guidance](httpswww.adsabs.harvard.edu/abs_doc/help_pages/data.html).
A candidate should be tested as a bibcode identifier in a bibliographic service; a locally chosen BibTeX citation key does not become a bibcode merely by resembling one.
A DOI, arXiv identifier or ADS page URL may lead to related material, but each must remain distinguishable from the exact bibcode token.
Scope
+ Exact bibcode text and its extraction from citations, URLs or bibliographic records
+ Interpretation of positional components under an evidenced bibcode convention
+ Evidence that a bibliographic service recognises and assigns the code
+ Agreement between the resolved record and the intended citation
+ Service-reported canonical, alternative or superseding identifier relationships
+ Permissible lookup, preservation, linking and correction actions
- The publication's scientific claims, methods and evidential quality
- Complete bibliographic descriptions and citation-style rendering
- Author identity and researcher identifier management
- Journal identity, editorial policy and publication schedules
- DOI registration and arXiv identifier lifecycle
- Full-text hosting, access rights and licensing
Characteristics
- Exact identifier text
- Literal character sequence, preserving case, punctuation and padding Prevents extraction or normalisation from silently changing the identifier.
- Token length
- Characters; conventional bibcode length is 19 Detects truncation, attached punctuation and incomplete search patterns without claiming that length proves validity.
- Representation assessment
- Unexamined | conventional form | documented exception | malformed | uncertain Separates ordinary parsing from exceptions requiring additional evidence.
- Component interpretation
- Year, publication abbreviation, volume-or-type, qualifier, locator and author initial; each interpreted, uninterpreted or disputed Supports citation comparison while making incomplete decoding visible.
- Assignment evidence
- Constructed candidate | externally asserted | service-confirmed | disputed Distinguishes a plausible generated string from a recognised identifier.
- Resolution observation
- Unchecked | matched | no match | redirected | ambiguous response | service failure; qualified by service and observation time Prevents a temporary lookup failure from being treated as proof of an invalid bibcode.
- Identified bibliographic record
- Service-qualified record reference with retrieval evidence Anchors interpretation to the actual record returned.
- Citation agreement
- Uncompared | consistent | partially consistent | conflicting | insufficient metadata A resolving bibcode can still identify the wrong item for the user's citation.
- Identifier relationship
- Canonical code | alternative code | reported replacement | DOI mapping | arXiv mapping | related-item link; each with evidence Supports linking without collapsing distinct publication versions or related records.
Also called
Where this came from
wikidata · CC0 1.0
Drafted structure
Bundle to layer to finding to question, as the second pass will find it: 5 bundles · 10 layers · 10 findings · 20 questions.
Token recognition Establish what exact string is being offered as a bibcode and whether it is complete.
Bibcode padding and positional structure make apparently harmless text cleanup capable of changing identification.
Capture context
Separate the identifier token from its surrounding representation.
Literal token
Record the supplied characters and how they were extracted.
- What exact bibcode text was supplied, and did it come from a record field, citation, URL or user entry? provenance
- Which surrounding characters or URL encodings were removed or decoded, and can the original representation be recovered? action
Form assessment
Distinguish complete conventional codes from malformed tokens and lookup patterns.
Complete code or pattern
Assess token length and positional plausibility without equating syntax with registration.
- How many characters remain after extraction, and do they fit the documented positional layout? measurement
- Is this a complete identifier, a partial bibcode search expression or a claimed exception requiring separate documentation? boundary
Bibliographic decoding Interpret the bibliographic clues embedded in the token without treating them as a complete citation.
Bibcode components assist recognition but require conventions that depend on publication type and locator usage.
Publication coordinates
Interpret year, publication abbreviation and volume-or-type positions.
Container and year
Associate positional clues with a supported publication context.
- Which year and publication abbreviation does the code express, and what authority identifies that abbreviation? definition
- Does the volume-position content represent a serial volume, a publication-type marker or an interpretation that remains unresolved? boundary
Locator and initial
Interpret qualifier, locator and author-initial positions conservatively.
Special-position semantics
Identify when locator-related characters or author initials need publication-specific interpretation.
- What documented rule explains the qualifier and locator characters for this publication, including any non-page usage? definition
- Does the final initial agree with the record's first author under the applicable convention, and what uncertainty prevents a reliable comparison? boundary
Assignment and resolution Establish recognition by a named bibliographic service and distinguish lookup outcomes.
A bibcode generated from citation details can look correct without being assigned to the intended record.
Assignment provenance
Identify the evidence behind the claim that the string is an assigned bibcode.
Recognition evidence
Separate service-returned identifiers from copied assertions and locally constructed candidates.
- Was this bibcode returned by a bibliographic service, copied from a secondary citation or constructed from metadata? provenance
- What retrieved record or response establishes that the service recognises this exact code? provenance
Lookup state
Record what happened when the exact bibcode was queried.
Observed resolution
Preserve service identity, observation time and returned identifier alongside the lookup outcome.
- Which service was queried, when, and did it return a matching record, another code, no match or an operational error? measurement
- Does the observed result justify using the code, retrying the lookup or investigating an identifier mismatch? action
Referent and equivalence Determine whether the bibcode identifies the intended bibliographic item and which identifier relationships are supported.
Successful resolution alone cannot establish citation correctness or equivalence between publication versions.
Citation match
Compare the retrieved item with the citation the agent is trying to identify.
Intended-item agreement
Assess agreement using record metadata beyond the abbreviated clues in the code.
- Do title, authors, year, publication and locator in the returned record agree with the intended citation? boundary
- Could the returned item instead be an erratum, abstract, proceedings contribution or another version requiring a distinct link? boundary
Identifier crosswalks
Qualify links to other bibcodes, DOIs and arXiv identifiers.
Supported equivalence
Record the asserted relationship and its source before using identifiers interchangeably.
- Which source explicitly connects this bibcode to another identifier, and what relationship does that source assert? provenance
- Does the evidence support the same bibliographic record, another version of the work or merely related material? boundary
Preservation and correction Guide reuse and repair of bibcodes while retaining the evidence behind earlier references.
An agent must avoid turning a suggested repair or metadata-derived candidate into an unsupported identifier assignment.
Canonical code handling
Manage differences between supplied codes and service-preferred identifiers.
Reported code relationship
Treat canonicalisation or replacement as an evidenced service assertion.
- Does the service explicitly identify a preferred, alternative or replacement bibcode, or has the agent only inferred one? provenance
- If a preferred code is adopted, how will the original supplied code and the supporting relationship evidence remain traceable? action
Repair and reuse
Choose actions appropriate to malformed, unresolved or conflicting identifiers.
Evidence-bounded action
Permit verified linking and supported repair while keeping unconfirmed candidates explicitly provisional.
- Can metadata search establish a verified correction, or must the candidate remain unresolved pending review? action
- Before placing the bibcode in a citation, request or URL, have its exact characters been preserved and any transport encoding kept separate from the identifier? action
What the second pass must settle
- Which current ADS, SIMBAD and NED conventions differ, and which service governs interpretation when they disagree?
- What documented rules cover article numbers, extended locators, non-journal items and historical exceptions to conventional positional decoding?
- How do the relevant services expose canonical and alternative bibcodes, and what guarantees apply after record corrections, merges or removals?
- How are collisions, unavailable author names and non-Latin or accented surnames handled under current assignment rules?
- What evidence is sufficient to distinguish a temporarily unresolved bibcode from an unassigned candidate, and how should that evidence expire?