Analysis assets
Version scripts, notebooks, datasets, environment files, reports and generated outputs.
Reproducible analytics
Connect methods, code, data and outputs in one evidence trail.
Register R, Python, notebook and external-runtime assets; preserve input checksums, environments, parameters, outputs, logs, reviews and portable reproducibility manifests.
Available · isolated worker optional
Core capabilities
Designed as part of the platform—not as an isolated feature.
Execution evidence
Record runtime, parameters, seeds, status, logs, input/output links and integrity checksums.
Isolated worker
Queue R and Python jobs for disposable no-network containers with CPU, memory, process and timeout limits.
Human review
Require researcher review and preserve final approval state before results are relied upon.
How it works
A controlled workflow from setup to review.
- 01
Define an analysis workspace and approved runtime.
- 02
Attach governed inputs and code.
- 03
Execute externally or through the isolated worker.
- 04
Review logs, diagnostics, outputs and assumptions.
- 05
Export a reproducibility manifest.
Frequently asked questions
Clear answers before activation.
Does the main PHP application execute arbitrary code?
No. Code execution is separated into an optional isolated worker using disposable containers and restrictive runtime controls.
Can external analyses still be recorded?
Yes. The workspace can register evidence from approved external systems without executing the code itself.
Next step
Make analysis reviewable.
Start a project and create a governed analysis workspace.