Skip to content

Trust & security

Every step on the record.

An agent you can’t audit is an agent you can’t trust. This is how Lumenar is built to keep a person in charge, and what we have and haven’t earned yet.

Approvals

Anything irreversible, like a refund, a merge or an email to a customer, waits at an approval gate. A named person approves or redirects it, and their name goes in the log.

Refund
Waits
Merge
Waits
Email to a customer
Waits

Append-only audit log

Every session leaves a trail: the plan, each step, each tool call, each approval. Entries are added, never edited, and the whole log exports.

The plan
Logged
Each tool call
Logged
Each approval
Logged

Open models, no lock-in

The session layer runs open models on vLLM and SGLang and speaks the OpenAI-compatible API. The session log belongs to no single model, so you can switch models mid-session.

vLLM
Runs on
SGLang
Runs on
API
OpenAI-compatible

How session logs are used

A session log is the record of one agent’s work. It is used to run that session, to hand it to another model and to show you what happened. In a pilot we agree in writing what is kept, where, and for how long.

Run the session
Yes
Hand to another model
Yes
Show what happened
Yes

Certifications

Planned, not claimed

Lumenar holds no certifications yet. These are the ones we are working towards, and they stay marked as planned until they are earned.

  • SOC 2 Type IIPlanned
  • ISO/IEC 42001Planned

Security contact

Found something?

Write to us with the details and how to reproduce it. A founder reads every report.

lumenar.dev@gmail.comEmail a security report