matthiasn/lotti

matthiasn/lotti

Agent

The persisted agent runtime: the agent kinds, their wake scheduling, memory, and the human review gates in front of every task mutation.

matthiasn/lotti

Agent

The append-only input event log, LLM-distilled summary checkpoints, the byte-stable prompt prefix invariant, state-as-projection, and fork healing.

matthiasn/lotti

Agent

The persisted agent runtime — five agent kinds, their startup wiring, lifecycle, and the boundary against the AI inference stack.

matthiasn/lotti

Agent

The agent.sqlite entity and link model, bulk-read chunking, and exactly what leaves the device.

matthiasn/lotti

Agent

The digest-shaped project agent that resists waking on every linked-task edit, and the leaner event agent that writes recaps under a hard human-authorship invariant.

matthiasn/lotti

Agent

A pure, deterministic fold over an event set — proving that projected state is the same regardless of arrival order or branching.

matthiasn/lotti

Agent

The primary agent workflow — inference setup resolution, the automation switch, evidence-first execution, tool policy, and the proposal/confirmation loop.

matthiasn/lotti

Agent

What an agent does (template skills) versus who it is (soul personality), and the ritual loop that evolves both from user feedback.

matthiasn/lotti

Agent

The AI summary card and its proposal choreography, the internals panel, the settings tabs, and the sidebar wake queue.

matthiasn/lotti

Agent

How a local change becomes an agent wake — subscription matching, run-key dedupe, workspace partitioning, bounded concurrency — and the three failure modes the design defends against.