About SkeleTrace

What we build

SkeleTrace brings three instruments into one program: an engine for entities, relations and measurements over time; a case notebook for public records and search results; and a landscape that reads traffic for structure. All three write plain, documented JSON, and all of it opens in a viewer that runs entirely in the browser.

The software is written in Rust and runs from the command line on Windows and Linux. The case notebook shares its file format with the Python OSINTropy tool, so either can open the other’s cases.

How we work

The design follows a small number of rules, and they are not relaxed for convenience.

• Provenance first. Every observation records where it came from and when. Nothing is merged unless a person decides it should be, and the reason is kept with the link.
• Restraint in what we claim. A matching name is not an identity, and a statistical anomaly is not an attack. The software says what the evidence shows and stops there.
• Published interfaces only. Connectors use documented public APIs, pace their requests and identify themselves in every request.
• Privacy by construction. The viewer is not permitted to make network requests. Credentials stay in environment variables and never appear in records, reports or error messages.
• Limits stated plainly. Where something is untested or uncertain, the documentation says so.

Licence

SkeleTrace is released under the MIT licence.

Contact

For questions about the software, the data formats, licensing or security, write to [email protected].

When reporting a security issue, please include enough detail to reproduce it, and say whether you would like to be credited.

This website

These pages set no cookies and load no trackers, advertising or third-party scripts. The viewer is a single page that runs in your browser: the files you open in it are read on your computer and are never uploaded.