Software for documenting, registering, and accelerating scientific research

A protocol for documenting your existing workflows in a self-hosted decentralized knowledge graph. Every claim in a Composable PDF is automatically and immutably linked to its originating code, data, and metadata.

Python
LaTeX
Composable PDF

Total Attestations

14,122

Unique Attesters

2,150

Most Common Claim Types

Based on all attestations in the ledger.

  • Computational Verification
    7,799
  • External Data Import
    2,451
  • Private Computation Proof
    863
  • Peer Review
    3,009

Transforming Scientific Workflows

Composable Science provides tangible solutions for the entire research ecosystem.

Data Integrity for University Labs

Secure your lab’s foundational datasets. With a single attestation, create a canonical, immutable version of any dataset, ensuring every collaborator builds from a single source of truth.

Verifiable Regulatory Pipelines

Revolutionize regulatory audits for clinical trials. Pre-register analysis code before data is unblinded, creating a cryptographically secure, timestamped audit trail that is impossible to tamper with.

Reproducibility for Researchers

Publish papers with embedded, executable proof. A single ID links to your entire methodology, allowing reviewers and readers to reproduce your results, byte-for-byte, with a single command.

De-Risking Deep Tech Investment

Move beyond trust to cryptographic truth. Provide investors with a real-time, un-hackable audit trail of your research, drastically reducing due diligence friction and building deep confidence.

Reinventing the Scientific Journal

Power your entire publishing workflow—from submission and automated review to beautiful, living articles—on a single, verifiable platform.

Document-Driven Workflows

Use the power of the written word for organizational alignment. Treat proposals, plans, and reports as living documents verifiably linked to the work itself, providing unprecedented organizational clarity.

Explore the Ledger

Attested by did:key:z6Mkk7yqnGvTj4e9... on July 29, 2024

This attestation successfully replicates the protein structure prediction for the SARS-CoV-2 spike protein receptor-binding domain using AlphaFold 2, confirming the reproducibility of the original computation.

Attested by did:key:z6MkqE2yqXnU7w9p... on July 28, 2024

A climate model simulation based on the CESM2 model under the SSP2-4.5 scenario projects a global mean temperature anomaly of 1.5°C relative to pre-industrial levels by the year 2040.

Attested by did:key:z6MkfT3zNoP9q8z9... on July 27, 2024

No summary available for this attestation.

Attested by did:key:z6MkhG5pLbV6z3fW... on July 26, 2024

No summary available for this attestation.

Attested by did:key:z6MkaaaaaaV6z3fW... on July 25, 2024

A high-trust assertion from a curator for the ATLAS open dataset from CERN. This makes the data available as a canonical artifact for other verifiable computations.

Attested by did:key:z6MkSharmaLabPri... on July 30, 2024

A privacy-preserving computation over sensitive clinical data. This attestation provides a public result (mean recovery time) backed by a Zero-Knowledge Proof, ensuring computational integrity without exposing patient information.

Attested by did:key:z6MkaaaaaaV6z3fW... on January 15, 2025

Revealed clinical trial data corresponding to the private computation claim attestation-006.

Attested by did:key:z6MkAuditCommitt... on January 20, 2025

No summary available for this attestation.

Attested by did:key:z6MkResearcherPr... on June 1, 2024

This attestation serves as a timestamped commitment to a specific version of an analysis script before the trial data was unblinded.

Attested by did:key:z6MkDataCuratorF... on July 15, 2024

Import of the unblinded dataset for drug trial XYZ.

Attested by did:key:z6MkResearcherPr... on July 16, 2024

This verifiable computation runs the pre-registered analysis script on the subsequently revealed and randomized trial data, confirming the study's primary endpoint.

Attested by did:key:z6MkBotsAreCoolR... on July 15, 2024

A trusted bot has deterministically randomized the participant IDs into treatment and control groups to ensure unbiased analysis.