VESPERA

Thesis

A local-first AI orchestration runtime for owned context and human-approved workflows.

Problem

Modern AI workflows fragment context, decisions, assets, and approvals across disconnected tools and sessions.

System

VESPERA treats AI tools as replaceable workers while keeping context, routing, approvals, and project memory under user control.

Implemented in the recorded prototype

• Local-first positioning and workflow philosophy
• Open-source / MIT public project positioning
• Human approval philosophy for higher-risk workflow levels
• Context, routing, and workflow control narrative
• Public repository and public runtime landing page

Status

v0.2-alpha

Open-source local-first AI orchestration runtime in alpha development.

Planned, not yet implemented

• Evaluate feature availability against each repository release
• Replace preview capture as runtime screens mature

Links

Evidence in the project record

• Live page identifies v0.2-alpha, MIT, local-first, zero telemetry, and human approval for Level 2+.
• Repository describes VESPERA as a local-first AI operations kit.