Launch
Give new research the best possible path into the scientific ecosystem.
Most work that goes unread does not lose an argument — it loses a link, a license field, or a structured abstract. Trace Launch handles the unglamorous layer: availability, identifiers, machine readability, and distribution that is attributable to a named human being.
Distribution rules, without exception
No fake accounts. No synthetic endorsements. No hidden promotion.
Every post is made from a disclosed account with the author or institution named. Nothing is published without a human approval step, and the approval log is retained.
Availability
A paper that cannot be reached cannot be retrieved.
Availability is checked as an outside reader and as a crawler, because the two often see different pages. Findings are reported as observations with a date, not as guarantees about future indexing.
- Resolvable identifier
- The DOI resolves, and the landing page it resolves to is reachable without a login.
- Access status
- Open, green, closed, or unknown — recorded as observed, not as intended.
- Repository deposit
- Whether an accepted manuscript exists in a repository the engines can actually read.
- License clarity
- Whether reuse terms are machine-readable, which determines what may be indexed at all.
- Full-text reachability
- Whether crawlers receive the full text or only an abstract behind an interstitial.
- Persistence
- Archiving arrangements, so a link that works today still works in five years.
Metadata
The fields that decide who ever sees it.
Metadata completeness is checked field by field against what indexers and retrieval systems actually consume. Each gap is reported with the consequence it has, so you can decide whether it is worth fixing.
Checked fields
Completeness review
- TitleSpecific enough to answer a question, not only to describe a topic.
- AbstractStructured, with population, comparator, and outcome stated plainly.
- Keywords and MeSHMapped to controlled vocabulary, not invented terms.
- Author identifiersORCID on every author, not only the corresponding one.
- Funding and conflictsMachine-readable, in the fields consumers actually parse.
- RegistrationTrial or protocol registration linked from the record itself.
- Data and codePersistent links with their own identifiers and licenses.
- Related versionsPreprint, accepted manuscript, and version of record connected.
Machine readability
Write it once for people, once for parsers.
Structured records are emitted in the formats retrieval systems already read, and validated before they are published. Nothing here changes the science; it changes whether a machine can tell what the science says.
Demo workspace
Structured record for the fictional TRACE-101 trial
Identifiers use the 10.5555 demonstration prefix and resolve to nothing.
- Identifier
- 10.5555/trace101.2026
- Registration
- Linked from the record, not only the PDF
- Population
- Adults with the fictional condition studied
- Intervention
- The fictional investigational agent studied
- Primary outcome
- Stated in a parseable field, with the timepoint
- License
- Machine-readable reuse terms
Fictional record created for demonstration. It is not a real trial and must not be cited.
Author and institution assets
The record around the paper is part of the paper.
Drafted in the workspace, reviewed by a named author, and published under that person's or institution's name.
- Plain-language summary
- One paragraph a non-specialist can read without acquiring a misconception. Drafted for review, never auto-published.
- Author profile alignment
- Consistent name form, affiliation, and identifier across the record, so the work aggregates to the right person.
- Institutional record
- Repository entry and institutional page carrying the same structured metadata as the publisher record.
- Figure and result descriptions
- Alt text and captions that state what the figure shows, which is also what makes it machine-readable.
Transparent community participation
Participate as yourself, or do not participate.
Sharing a finding in a community where researchers are already discussing the question is legitimate. Doing it from an account that hides who is speaking is not. There is no version of this product where the second one is available.
Never
Out of scope, permanently
- Fake or undisclosed accounts
- Synthetic endorsements, reviews, or comments
- Undisclosed paid promotion
- Vote, metric, or download manipulation
- Automated posting without a named human approval
No fake accounts. No synthetic endorsements. No hidden promotion.
Approval workflow
Nothing leaves the workspace unreviewed.
Regulated workspaces add a compliance step and can block publishing entirely until it is cleared. The log is exportable for audit.
01
Draft
Generated or written in the workspace, marked as unapproved.
02
Author review
A named author edits and signs off. No silent acceptance.
03
Compliance review
Required where the workspace is set to a regulated mode.
04
Publish
Posted from a named, disclosed account with the affiliation stated.
05
Log
Retained with who approved what, when, and in which version.
After publication, keep measuring
Availability degrades quietly. Links rot, repositories migrate, and access status changes without anyone being told. Publication checks are re-run on a schedule and recorded as history, alongside retrieval and citation measurement. Changes are annotated as occurring after an event, never as caused by it.