Practice-born. Research-backed. Formally generalized.

Systematizing knowledge for science, engineering, and public reasoning.

Boundary First Labs is an applied systems research laboratory studying scientific software modeling: how representations become executable, testable, and consequential.

Software Before Code is the center doctrine. Boundary-First Engineering turns it into practice; the Lab's research, products, and institutional machinery test how state, transition, invariants, authority, evidence, and repair survive real execution.

Explore Applied Work Read the Research

Lab snapshot

UpdatedSep 29, 2026

96Kfiles

621durable registries

262Ksemantic units

HOW THE LAB IS ORGANIZEDThe numbers sit on top of a working institutional machine.

Public interface2026-10-02

The public website was brought into line with the Lab's current doctrine.Open affected surface

Institution2026-10-01

The Lab's internal priority shifted from adding components to closing complete loops.Open affected surface

Research2026-10-01

The Lab added an executable check against double-counting dependent evidence.Open affected surface

CHOOSE YOUR OWN PATH

Start with why you came.

Open all audience paths

OUR APPROACH

What the Lab does.

OPERATING BRAID

Three lineages become one method.

STEWARDSHIP

What succeeds still has to be cared for.

CONSULTING / APPLIED WORK

Bring one system that is expensive to misunderstand.

Three bounded entry points grounded in scientific software modeling: systems architecture, agency / AI governance, and knowledge / representation infrastructure.

Explore Applied Work Start a conversation

Systems / Architecture Review

Reconstruct the system · expose defects and invariants · compare repair / migration paths

Agency / AI Governance Audit

Map decision rights · define human / agent gates · repair authority and provenance gaps

Knowledge / Representation Infrastructure Diagnostic

Map source / claim / evidence state · find representation loss · design durable handoff infrastructure

Systems / Architecture Review is the current first-engagement focus. The other two offers remain available when authority/governance or knowledge/provenance is the primary problem.

FEATURED WORK

Things we are building, testing, and publishing.

Concrete examples first. The labels simply tell you what kind of work each one is.

METHOD + ENGINEERING DOCTRINE

Software Before Code

A scientific software-modeling doctrine for constructing executable models that remain adequate under real-world consequence.

ACTIVE-BUILD PRODUCT FAMILY

Projectr

Public knowledge infrastructure for durable, source-linked knowledge; YouTube Knowledge Explorer is the current bounded implementation.

RESEARCH INFRASTRUCTURE

Corpus Forge

Executable scientific and research-operations machinery for bounded work, evidence, criticism, verification, repair, and promotion.

RESEARCH PRODUCT

Boundary-First Chess

A book-length teaching asset and developed pedagogy for making structural change on the board more legible.

THE LAB IN MOTION

See the current state. Choose a way in.

NOW / ROADMAPOCTOBER 2026 · CURRENT CYCLE

Externalize the clarified doctrine, test it, and close the obvious gaps.

Observe → bound → represent → execute → witness → criticize → repair → transfer.

01Public institutional interface02BFL-native external evidence03Representational laboratory program04Bounded publications + criticism05Relationships, funding + distribution06One coherent body of work

Open Now / Roadmap

COLLABORATIONReview · test · build · distribute · transfer

Bring a real problem, capability, audience, or resource.

Explore Collaboration

FUNDING / CAPITALIZATIONRunway · earned services · product capital · research capital · evidence-gated credit

Capitalize the conversion engine, not the theory.

See the Funding model

OUR STANCE

Working posture

Sober, inspectable, and criticism-friendly. The Lab is built to absorb counterexamples, revise its machinery, preserve human agency, and leave useful capability behind.

“Better systems for a more legible world.”

INSPECTABILITY

Show enough of the machinery to be challenged.

Claims should travel with their representations, evidence, uncertainty, dependencies, and repair paths so another person can inspect more than the conclusion.

Why representation and repair matter

CONTESTABILITY

Criticism should be able to change institutional state.

Counterexamples, missing distinctions, implementation defects, accessibility failures, and stronger evidence should have a route into revision rather than disappearing into a generic inbox.

See the Open Lab participation model

CAPABILITY TRANSFER

Useful work should leave more capability behind.

The preferred outcome is not permanent dependence on the Lab. It is clearer models, reusable tools, repair paths, documented processes, and handoff to people who can carry the work forward.

Read the stewardship posture

AI GOVERNANCE

Forge what helps. Certify what acts. Forbid what dominates.

AI makes the agency question concrete: bounded assistance can accelerate, while consequential delegated authority needs explicit boundaries, contestability, repair, and accountable ownership.

How we govern AI