A persona layer for a world about to be very full of agents.

What gosum is

gosum is an open standard for personas — the layered, composable narrative artefacts that give an AI agent a voice, a memory, and a sense of where it is from. It is also a runtime: a small, governable engine that loads a persona file and behaves accordingly.

A persona is not a system prompt. It is not a configuration. It is a small library of voice, lineage, and memory, signed and auditable, written in a vocabulary anyone can read.

The standard is free and open. The runtime is open source. What you pay for, eventually, is help building good personas — and the tooling to govern them at scale.

Why it exists

In the coming decade, most software you talk to will need a voice. Most robots will need a small story about themselves. Today, that voice is improvised — a long system prompt, tweaked weekly, owned by no one in particular, drifting under model updates.

gosum proposes the opposite: that personas are first-class artefacts, sourced from human literary tradition, governed like any other sensitive system, and portable across runtimes. Less improvisation. More craft.

Who’s behind it

I’m a writer-engineer. The work is hosted by the Folklore Archive Institute and a small Romanian non-profit (Asociația) that holds the IP for the public-domain literary lineage. The institute provides the editorial review for new lineage entries.

Where it’s going

v.01 is the public simulator, the schema validator, persona signing, and the first governance layer. After that, the runtime opens to third-party hosts. See the architecture →

“In the coming years, the robots around us will not just perform tasks; they will tell stories, share jokes, offer comfort — each in their own style.” liviu pălălău · 2013 · ipv6 forum