limina

What problem does it solve?

It solves the problem of turning an open-ended research effort into a structured, auditable workflow that produces evidence rather than guesses.

Core Features & Use Cases

• Persistent evidence loop in kb/: Maintains durable artifacts across long runs, linking hypotheses to experiments and experiments to findings.
• Hypothesis-driven research workflow (H -> E -> F): Breaks the mission into research questions, designs experiments, challenges assumptions, and iterates until success criteria are met or approaches are exhausted.
• Escalation with traceable reasoning: If the agent gets blocked, it asks you instead of guessing, while keeping a full trail of decisions and outputs for review.

Quick Start

Tell your AI to set up a new Limina project in a folder named limina-research and initialize the mission brief using your objective, context/baseline, success criteria, resources/boundaries, and blocked stop condition.