Mycelia · engineering preview

Research,
in reusable parts.

Research moves before the paper does.
Keep the methods, data, code and discoveries connected as they happen — and let credit follow their reuse.

Private by defaultLocal AI, with human reviewOpen records, exact versions

Connected, credited, inspectable

Follow the work.

Each point is an actual public version in this catalogue. Select a part to see where it leads.

0weighted reuseFrom the public event ledger

Local sandbox. Research examples and verification are fictional. Imported metadata identifies its real external source; no scientific effect is established by this demo.

Live catalogue read · refreshes every 30 seconds

A part worth following

Methods connect. Evidence accumulates.

Move through the network with your keyboard, or choose the list. Every link opens the exact published version.

◇ Protocol
How the work was done
▦ Dataset
What was recorded
△ Claim / ▽ Null
What the evidence supports
✦ Verification
Who checked the exact version
Graph shows up to 24 public parts. The list shows up to 60. Private drafts are excluded.

From the bench to the next breakthrough

Less retyping.
More reusable research.

  1. 01

    Drop

    Bring files into your private lab memory. Local tools draft metadata; you decide what to keep.

  2. 02

    Link

    Connect the dataset to its protocol, the claim to its evidence, and the contribution to its person.

  3. 03

    Check

    See what each trust rung requires. Contributors sign off, and reproduction links stay visible.

  4. 04

    Reuse

    Cite an exact version, follow the evidence and see where reuse earns simulated credit.

Two rules. A clear record.

Humans approve.
Evidence stays attached.

Every contributor signs off. The data and code needed to reproduce a claim are linked, or the record explains why they are not available.

AI suggestions carry their model and provenance. A machine-generated review indicator is not peer review. Published versions remain signed and immutable; corrections create a new version.

Read the trust and privacy boundaries →

Start where the work happens

A useful place for your part.

For researchers

Your lab’s memory,
before publication.

Capture files, find earlier attempts and turn a reviewed draft into a linked research part.

Set up your workspace →
For labs

Trace the run,
keep the decision.

Inspect protocols, signed simulated runs and the approval behind every loop iteration.

Explore the lab workflow →
For funders

Fund verification.
Follow reuse.

Set a trust requirement, inspect gaps and reconcile simulated bounties and dividends.

Explore the funder workflow →

Open by design

The record belongs
with the research.

Import repository metadata. Export signed records, RO-Crate, nanopublications or JATS. Use REST and MCP to follow exact versions.

See formats and integrations ↗

This is a local prototype with fictional partners. No live service, real payout or production DOI is claimed.