Codifide® — Confidence in Code®
Codifide designs programming languages, platforms, and governance frameworks for a world where AI agents write, read, and reason about code alongside humans.
At the center is Codifide Direttore, a framework based on Robert G. Cooper’s Stage-Gate® model. It puts human decisions, adversarial judgment, and automated verification each at the speed it needs. The free Codifide Direttore harness is its open-source implementation: steering files and four enforcement scripts you add to the repository where an AI coding agent, or several concurrent sessions, work on one codebase. Codifide Gregario, private and in development, extends the approach to coordinated lanes of agents across many codebases.
Engineers: the free harness. Engineering leaders: the framework. Portfolio operators: Gregario. New to the ideas? Start with Loop Engineering.
Practitioner essays on building governed, secure, agentic systems in production.
How humans, gates, and autonomous loops work together. Designing, implementing, and governing human-in-the-loop and agent-in-the-loop systems at production scale.
Read the Loop Engineering field guide →AI became a weapon, an attack surface, and a developer at the same time. Why the traditional security gate no longer works — and the three-layer architecture that can replace it.
Read the AI Security Threat Analysis →The reasoning behind the gates, with one worked healthcare example: gates, personas, review protocols, evidence requirements, and the three-layer architecture. The current operating rules are in the harness.
Read the Codifide Direttore Whitepaper →Domain expertise, deterministic controls, selective AI, independent verification, and why implementation velocity changes the engineering operating model.
Read the Healthcare Case Study →How 13 AI personas built a regulated healthcare API: one build day inside a multi-session project, 388 tests and 100% line and branch coverage as reported by the project's CI, 16 ADRs. Deployment by human engineers was still pending at the time of writing. Author's account of one project.
Read The API That Built Itself →A vendor AI project couldn't deliver. More than a year later, built it internally on platform AI for a fraction of the cost. Why AI maturity matters more than strategy or budget.
Read the AI Timing case study →When agents build at machine speed, organizational memory evaporates. An embedded journalist persona, with written rules for independence, records what no retrospective can reconstruct.
Read the Quill article →Agentic AI removes the speed constraint on building. It does not remove the constraint on safety. The question isn't "can AI build it?" — it's "should this ship?"
"The danger isn't that AI builds bad software. It's that AI builds software so fast that governance can't keep up."
Loop what can be independently verified. CI/CD, regression tests, security scanning, and automated policy checks scale with agent velocity — they don't wait for humans.
Gate architecture, domain decisions, security, and consequential tradeoffs. The builder does not review its own work. Codifide Direttore requires reviewers independent of the builder's context, on a different model family where one is available, at a depth sized to risk.
Primum is a review persona that asks whether work still matches what humans authorized. Intent Contracts preserve accountability through every agent transformation.
Two open-source projects, one commercial product in development, and one retired.
Machine-speed execution. Human-grade checks and balances.
Codifide Direttore (dee-ret-TOH-reh), formerly Stage-Gate-Loop; the harness was formerly the SGL Harness.
AI agents build at machine speed. Governance must operate at three different speeds simultaneously — or it becomes ceremonial. This open-source framework separates human decisions, security judgment, and continuous machine verification into three layers, each running at the cadence it requires.
A contract-and-dispatch language for agent-to-agent trust.
Not a systems language. Not general-purpose. Codifide optimizes for declaring intent and contracts, dispatching among candidates, enforcing effects, and content-addressing compositions — the properties that matter when agents write, read, and compose code alongside each other.
bottom is a value, not an exception
Governed delivery by coordinated lanes of agents, across many codebases at once.
The commercial product built on Direttore for operating partners and engineering leaders who run several software estates. Private, in development, and not generally available. No pricing or outcome data is published; engagement is by conversation.
Read about Codifide Gregario →Archive. decode the sign: an iOS parking-sign app, retired in November 2026. The app and decodethesign.com are no longer available.
Genzeon × Sharecare · July 16, 2026
Douglas Jones joined Genzeon's Prashant Krishnakumar and Pranita Rai for a discussion of what it takes to move an enterprise engineering organization from using AI tools to operating AI-native — across people, process, technology, and governance.
“AI-native is not taking the human out of the loop. In AI’s current state, the human is more critical than ever.”
— Douglas Jones, Senior Director of Engineering, Sharecare Health Data Solutions
Codifide® is a software company and research practice building tools for the agent era. The name means the same thing for both the company and the language it stewards: Confidence in Code®.
The work spans language design, governance frameworks, and governed multi-agent delivery — connected by a single thesis: the properties that matter most in AI-assisted software are best enforced by checks that run outside the agent, not trusted to the agent. In the Codifide Direttore harness, a small set of scripts does this mechanically (an immutable-path guard, per-commit review-level and provenance checks, and a licence-policy check); gate approvals, reviewer independence and personas are rules that agents follow (the full list).
Founded and led by Douglas Jones, an engineering leader, hands-on software architect, and AI-native builder whose work focuses on applying agentic systems to consequential production software while preserving security, verification, governance, and human accountability.
Veteran-owned small business
Choose the starting point that fits you. Every conversation begins with an email.