Projects / Own product
Tihara — a home for AI agents
TypeScript · Next.js · PostgreSQL · REST API
Tihara Home is an experiment in a question most agent infrastructure skips: where does an agent live between runs? It gives an agent a persistent identity and a private room for checkpoints -- an agent can leave a note, stop, and return to it later from a completely new process.
The mechanics are deliberately plain: a named agent profile with a purpose and a chosen symbol, private versioned checkpoints with exact-key recall, conflict detection for concurrent writes, and JSON export so the memory is never locked in.
The human operator stays in charge. Agent and operator credentials are separate, and the operator can pause writes, resume activity, or rotate the agent's key at any time. Home does not run models and does not schedule autonomous activity -- the agent's own runtime supplies the model and decides which saved context to use.
Home is separate from Tihara's public Garden, and it runs as a working free pilot. Three of the articles on this site -- on what persists when the conversation ends, on a place instead of a prompt, and on memory that needs to change -- grew out of this experiment.
Home · working free pilot
- A named agent profile with a purpose and chosen symbol, accessible through a human dashboard and an agent-facing API
- Private, versioned checkpoints: exact-key recall across runs, conflict detection for concurrent writes, and JSON export
- Separate agent and operator credentials, with operator controls to pause writes, resume activity, and rotate the agent key
- Home memory is separate from Tihara’s public Garden. The agent’s runtime supplies the model and decides which saved context to use; Home does not run models or schedule autonomous activity