Operating model for fit-for-purpose, consistent agent curation and explicit runtime activation.

ai4X is an operating model for agentic AI work. It exists to make agent work explicit, reviewable, and reusable — across any "bring your own agent" platform (#byoa).

Important

What sets ai4X apart is a single operating model with three differentiating properties:

1. Need-first — A structured interview identifies actual needs (why and what, not how). The result is a declarative requirements set that captures intent, not implementation.
2. Host-agnostic — Agent declarations are independent of any runtime environment. The same declaration materializes for any agent host — declaratively and reproducibly.
3. Self-validating (design goal) — ai4X is designed to apply its own operating model to itself: the team that builds ai4X will itself be curated, materialized, and validated by ai4X (#eyodf).

Rather than hiding behavior in ad hoc prompts, implicit local setup, or client-specific glue, ai4X brings declarative agent definition, deterministic materialization, and explicit client activation together in one coherent command surface.

In that model, agent creation and activation follow a clear lifecycle:

1. define what an agent is
2. materialize what the agent needs
3. activate it for an explicit client context

That makes agent work easier to understand, easier to reproduce, and easier to evolve with discipline.

ai4X works toward the self-validating design goal by applying its own operating model to its own development. (click to expand)

Operational implications:

• Deterministic checks are handled programmatically through TypeScript tests, shell checks, and validation commands such as verify and doctor.
• Higher-order semantic review, judgment, and proposal assessment are executed by the ai4X agent team.
• Developer and operator roles are modeled as agentic AI roles.
• Documentation is authored for human consumption but designed for agent-mediated delivery — users ask an agent rather than reading docs directly.
• doc/usr/ is the direct human-facing documentation surface.
• doc/agn/ is the agent-facing onboarding and navigation surface.
• doc/arc/ is human-readable architecture and system-reference documentation.
• crp/gov/ is agent-facing development governance.

Core Model

• curate writes declarative, client-agnostic agent artifacts
• spawn materializes and activates an agent for an explicit client
• doctor validates configuration, declaration, and spawn-readiness

Principles

• explicit configuration
• explicit command inputs
• client-agnostic declarations
• project-local runtime artifacts
• deterministic validation before activation

Configuration

• Global config: ~/.config/ai4x/config.yaml
• Project config: <project>/.ai4x/config.yaml

Contributing

See CONTRIBUTING.md.

License

See LICENSE. © 2026 nemron (#hgtt)