<!-- generated by scripts/generate-agent-docs.ts -->

> **Writing**
> Thinking about AI systems, deployment, and human-AI coordination. HALA pattern language for AI role allocation; agentic delivery with specs, supersession rules, and eval gates.
>
> Source: https://jasonstiltner.com/writing/

---

Frameworks and thinking about AI systems, deployment, and the organizational dynamics that shape what gets built.

## [When Production Disagrees with the Architecture](https://jasonstiltner.com/writing/when-production-disagrees/)

A caller answered two questions when our voice agent had asked one. That small production failure exposed a larger problem: evals can preserve counterexamples without preserving what they should teach us about architecture. Production evals → architectural assumptions → prospective evidence → earned autonomy.

[Reality-graded architecture loop](https://jasonstiltner.com/writing/when-production-disagrees/#architecture-loop) · [What exists today, and what is proposed](https://jasonstiltner.com/writing/when-production-disagrees/#act-ii-theory) · [References & intellectual lineage](https://jasonstiltner.com/writing/when-production-disagrees/#references)

## [The Archive That Cites Itself](https://jasonstiltner.com/writing/archive-that-cites-itself/)

Persistent AI advisors, the people they model, and the limits of “more context”—why a user model should preserve the history of its claims rather than a polished conclusion.

[Interactive mirror](https://jasonstiltner.com/writing/archive-that-cites-itself/#section-viii) · [Event-sourced state, applied](https://jasonstiltner.com/writing/archive-that-cites-itself/#section-x) · [References](https://jasonstiltner.com/writing/archive-that-cites-itself/#references)

## [HALA: Human-AI Layered Architecture](https://jasonstiltner.com/writing/hala/)

Pattern language for human-AI role allocation: 28 patterns across 5 layers. Inspired by observations from production AI work, where organizational dynamics—not technical capabilities—were the binding constraint. A design proposal; none of the patterns have been deployed.

[Interactive demo](https://jasonstiltner.com/writing/hala#demo) · [PDF](https://jasonstiltner.com/HALA_v2_Pattern_Language.pdf)

## [Agentic Delivery: Specs, Supersession, and Eval Gates](https://jasonstiltner.com/writing/agentic-delivery/)

How this site’s content was written by an AI coding agent under a written spec with supersession rules, a fact inventory, and eval gates—including the defect the process found in its own verification record.

[Process artifacts](https://jasonstiltner.com/process/README.md) · [Eval runs](https://jasonstiltner.com/process/evals/README.md)

## [Strategy Tournament](https://jasonstiltner.com/projects/strategy-tournament/)

An unimplemented design specification for a backtesting framework: how systematic strategies would be compared under realistic execution constraints—slippage, latency, partial fills—and robustness stress tests, so a strategy that only looks good against a frictionless backtest gets caught before capital does. Addresses evaluating adaptive systems when the evaluation itself can be gamed. No code was written and no repository exists.

[Why this page has no code behind it](https://jasonstiltner.com/projects/strategy-tournament#limitations)

[Atom feed](https://jasonstiltner.com/writing/feed.xml)

[jason@jasonstiltner.com](mailto:jason@jasonstiltner.com)
