I build agentic software that stays inspectable by its operator: clear state, explicit permissions, practical recovery paths, and tooling that can be understood after the magic wears off.
|
|
| System | Role | What it is for |
|---|---|---|
| Lilith CLI | creator / maintainer | Terminal-first AI workspace with durable missions, capability-scoped tooling, memory, orchestration and a Textual IDE. |
| Hermes Agent | fork / upstream contributor | Working fork of NousResearch/hermes-agent, focused on native Windows reliability and upstream fixes. |
| BrierStudios.com | designer / engineer | Public front door for dark-fantasy AI, software and creative-technology work. |
| Upstream PRs | contributor | Compatibility, testing and reliability patches across agent, memory and developer tooling projects. |
See Upstream Work for the contribution themes I tend to focus on.
Yggdrasil is the architecture I use to keep AI, tooling, interfaces, memory, creative systems and archives in explicit domains instead of letting everything collapse into one shapeless monolith. The operational core stays private by design; public-facing pieces are published when they are useful outside the local ecosystem.
BrierStudios is the studio identity behind the work. Its public home is brierstudios.com, while GitHub ownership is centralized here under BrierAinz.
|
|
|
Local-first by default. Clear boundaries over accidental complexity. Evidence before confidence. Recoverability before cleverness. Tools should amplify their operator without becoming opaque.


