Answer first
The tools section turns ModelBreeder theory into browser-local worksheets. Use the scorecards, simulators, router lab, lineage viewer, and release packet builder to reason about useful descendants before implementing a real model-breeding runtime.
Tools for constructive model-breeding decisions
The tools make the theory visible. They run in the browser, require no database, send no data to a remote form handler, and are designed as teaching surfaces that can become real project worksheets.
| Tool | What it teaches | Open it |
|---|---|---|
| Local AI Adoption Planner | Score a workflow for local-first, hybrid, or lab-first adoption. | Open |
| Fitness Scorecard Calculator | Compare a descendant with champion, specialist, challenger, and no-op choices. | Open |
| Local AI Opportunity Scorecard | Pick the first private, repeated, local-first workflow for a useful specialist. | Open |
| Population Simulator | Mutation, selection, carrying capacity, diversity, and retirement. | Open |
| Evolution Dashboard | Population state, novelty spread, role assignment, and next action. | Open |
| Lineage DAG Viewer | Parentage, operators, evidence, lifecycle state, and rollback. | Open |
| Router Policy Lab | Local specialist, adapter stack, cascade, ensemble, review, and no-op. | Open |
| Release Packet Builder | Copyable evidence packet for shadow, canary, promotion, rollback, or archive. | Open |
| Merge Recipe Sketchpad | Parent selection, operators, coefficients, and equal-budget evaluation. | Open |
| Adapter Stack Planner | Base model, adapter order, compatibility, and local runtime fit. | Open |
| Model Ecology Glossary Tool | Fast lookup of core vocabulary and where to go next. | Open |
Use the outputs as evidence starters
A scorecard or simulator result is not a final truth. It is a structured starting point for an evidence packet, architecture decision record, or lab note.
PROCEDURE use_browser_tool(tool, project_context)
assumptions <- RECORD_ASSUMPTIONS(project_context)
result <- tool.RUN_LOCALLY(project_context)
next_step <- SELECT_NEXT_EVIDENCE_ACTION(result)
RETURN CREATE_LAB_NOTE(assumptions, result, next_step)
END PROCEDURELocal AI adoption tools
Source reports used for this guide
These reports are preserved verbatim in the site archive. The guide above is an editorial synthesis and may narrow, qualify, or reorganize claims from the source material.