Public open-source research preview candidate · 0.3.0-alpha.1

From Analogy to a Testable Bridge

UVLM Governed AHA Lab Community Edition helps a human document whether a distant structural analogy preserves relations, names what does not transfer, generates a target prediction, and states the smallest test that could lower confidence.

  • MPL-2.0
  • CC BY 4.0 content
  • Local-first
  • No model call
  • No automatic donor selection
Not a truth engine. A reviewable bridge is still a candidate. The decisive question is whether the resulting experiment is useful and falsifiable.

What it does

  1. Accepts one local target and two to five explicit donor sources.
  2. Binds nodes and relations to exact source segments.
  3. Checks relation type, causal orientation, source-family independence, invariants, and disanalogies.
  4. Keeps epistemic posture, bridge quality, test value, and attention priority separate.
  5. Writes a closed deterministic evidence package and append-only human decision.

Downloads

Public source

Complete standalone source, tests, schemas, examples, licenses, SBOM, and docs.

SHA-256
e48ac229c43307e09feda323eed4bf2f696d4d1aaf6aae596695cab72a941d9a

Source ZIPChecksum

Exact review

Role-separated evidence, acceptance results, package checks, and limitations.

SHA-256
5f33abc2089b71b55a01cf617619860069407cf8e050e92845a6b5ed562623b7

Review ZIPChecksum

Companion manuscript

From Analogy to Governed Insight, v1.2.1 Open-Source Product Companion Edition. The 65-page design-science manuscript explains the larger governed AHA research program while distinguishing the working manual lab from capabilities that remain proposed or unproven.

SHA-256
6c96be42f9ddaa27dddb1576b74449f1112b71ff56443fa5dd47c0aac0ac2e65

Manuscript PDF Checksum

Known limitations

  • Graphs and mappings are human-declared.
  • No automatic discovery, calibrated truth probability, or baseline-superiority claim.
  • The browser interface is a loopback-only single-user research surface.
  • Fresh native-Windows reviewer execution remains open.
  • Human usefulness and accessibility require broader empirical study.

Read the companion website article · Support UVLM