Writing Hub
AI governance essays, reasoning systems notes, experiment logs, and technical writing across BioAI and engineering practice.
Project Topics
When the Memory Gate Met a Real Archive: What 90 Experiments Taught Us About Cheap LLM Slop
How to enforce data integrity against AI-generated slop using MICA. Explore a 11-step session-start validator that locks rules, playbooks, and contracts in code before code is ever touched.
The Two Problems No One Talks About in AI Agent Coding Pipelines
AI agent coding pipelines fail not because models are weak, but because verification is structurally broken. This article identifies four empirically documented failure mechanisms — agreement bias, latent entanglement, echoing, and right-for-wrong-reasons — and proposes a concrete architecture: hash-chained audit records, hybrid recurrence scoring, dynamic context budgets, and evidence-first review across three independent axes. Covers multi-agent pipeline design, agentic code review, blueprint indexing, and P0–P4 governance gates.
The README Was a Protocol. The Entrypoint Was Still Optional.
README-as-Protocol solved explicit invocation at the schema level. It did not solve entry control at the workflow level. This version adds the missing hierarchy: natural, guided, and forced activation.
Building a Deterministic Governance Kernel: Separating Custody from Truth
CGF separates domain truth from custody mechanics, turning AI governance from Markdown/YAML policy language into deterministic, inspectable artifacts.
The Difference Between a Harness and a Leash
A practical essay on why most AI 'harnesses' are still leashes: guides shape behavior, but only justified external measurement creates a real governance boundary.
The Next AI Moat May Not Be the Harness Alone: A Mathematically Governed Self-Calibrating Code-Review Layer
As AI harness patterns normalize, differentiation is shifting toward governed self-calibration and implementation fidelity. This piece explores how history-driven, bounded adaptation creates a new layer of defensible AI infrastructure — one that turns local code evolution into a competitive moat.
My AI Maintainer Kept Making Wrong Calls. So I Made It Report Its State Before Touching Anything.
Part 6 moves from landscape to operation. This is what MICA looks like when it is actually running inside a real maintenance workflow — session report, self-test, drift, invariants, and operator judgment.
Prompt → RAG → MCP → Agent → Harness, and What?
Why the next layer in AI may be governance infrastructure, not just better agents.
Everyone Was Talking About Context Engineering. Nobody Had Solved Governance.
Everyone Was Talking About Context Engineering. Nobody Had Solved Governance.
The Model Already Read the README. MICA v0.1.8 Made It a Protocol
v0.1.7 made scoring a contract with fail-closed gates. v0.1.8 recognized that README-first behavior could serve as invocation — and formalized it as a schema-level protocol. This article uses simplified examples to show how the invocation gap that had existed since v0.0.1 was finally closed
The Stake Was Governance Outside the Schema. MICA v0.1.5 Pulled It In
v0.1.0 through v0.1.4 made the schema more implementable. v0.1.5 was the first version to ask a different question — what if governance itself belongs inside the schema? Here is what that looked like, and what it still could not do.
The Schema Existed. The Model Had No Way to Know.
v0.0.1 proved that context could be structured. It did not prove that the structure could govern what shaped the session. Three failures — and why only one made the others meaningless.
Showing page 1 of 2 · 21 matching posts