CLARIXO Paths

Three entry paths for AI systems: LLM-side, User-side, and Development Platform.

Start with the path that matches your immediate system need. Use the LLM-side for runtime control, observability, approval, and governed execution. Use the User-side for behavior records, retained evidence, and formal review. Use the Development Platform when your next step is bounded AI product building through OC Platform and OC Workspace.

Explore LLM-side Explore User-side Explore Development Platform
This page is the side-selection layer. If you already chose the LLM-side, continue into Runtime Integration as the concrete start page for that path.
Why the paths are separated

Runtime control, user-facing evidence, and bounded development belong to different parts of the same system

CLARIXO separates provider-facing runtime control, user-facing evidence and review, and bounded development work because they sit at different points in the system and solve different operational problems.

LLM-side paths govern runtime execution, observability, approval, and governed handoff. User-side paths preserve what AI did for real users, then support later review on top of retained records. Development Platform paths give teams a dedicated OC-based build environment for structured implementation work inside CLARIXO.

Core Runtime API is not an evidence export API.

Evidence API is not a runtime control API.

Development Platform is not just another API surface. It is the bounded OC build path inside CLARIXO.

Choose your path

Start with the side that matches your immediate need

First decide whether your next need belongs to LLM-side runtime control, User-side evidence and later review, or Development Platform implementation work.

LLM-side

Runtime control and governed execution

Use this side when your next step is controlling how AI runs, how runtime signals are reviewed, and how execution is governed across provider-facing paths.

Where it sits
Application → CLARIXO Core Runtime → External AI Provider
Core Runtime API Routing, fallback, guard visibility, runtime memory, and runtime control between your application and external AI providers.
AgentOps · Runtime Review Layer Use AgentOps when you need run visibility, runtime reading, review context, and clearer operator understanding before runtime issues spread.
Approval · Governed Execution Layer Move into Approval when runtime review is no longer enough on its own and execution needs formal approval, block, escalation, and governed handoff.
Runtime Integration This is the concrete LLM-side start page. Continue here after choosing the LLM-side path.
Runtime docs Read the runtime contract, docs, and implementation materials after choosing the LLM-side path.
Start LLM-side Open Core Runtime API Read Runtime Docs
Development Platform

Bounded AI product building inside CLARIXO

Use this side when your next step is not only runtime control or evidence export, but structured product building, workspace continuity, and OC-based implementation work.

Where it sits
Team / Workspace → CLARIXO OC Platform → OC Workspace / Build Flow
OC Platform The bounded development platform for building AI products under CLARIXO architecture rather than entering only through runtime or evidence surfaces.
OC Workspace Use OC Workspace when you need the main working surface for task continuity, structured development flow, and implementation-side progress inside the platform.
Platform entry path Choose this when the immediate need is opening the dedicated development universe instead of starting from a pure API integration path.
Open OC Platform Open OC Workspace
User-side

Evidence first. Review access later.

This side starts with evidence export, then moves into more advanced review workflows when responsibility requirements become stronger.

Where it sits
AI System → CLARIXO Evidence API → Review / Audit / Case Reading
Evidence API A free, standard-installation evidence export path for user-facing AI behavior. Adopt this first to retain traceable records without changing the live runtime path.
Evidence Integration Path Use the start page, manual, and protocol to implement evidence export and close the first verification loop in a real system.
Audit Workspace A paid review workspace built on top of retained evidence records for grouped evidence reading, reviewer context, case review, and audit-ready export, with access provided through Responsibility rather than standalone purchase.
When to move up Move here when retained evidence is no longer enough on its own and your organization needs formal case handling, professional review discipline, and audit-oriented export workflows.
Which side should you start with?

Use the side decision that matches your current need

Start with the LLM-side if you need to:

  • Route across providers
  • Manage fallback behavior
  • Inspect runtime and guard behavior
  • Add approval or governed execution
  • Stabilize AI operation before it becomes a production problem

Start with the User-side if you need to:

  • Keep user-facing AI behavior records
  • Preserve traceable output evidence
  • Support review, audit, or dispute handling
  • Reconstruct incidents later
  • Add an evidence layer without changing your runtime path first

Start with the Development Platform if you need to:

  • Open a dedicated OC build environment
  • Work from OC Workspace as the main operating surface
  • Keep implementation continuity inside CLARIXO
  • Move beyond pure API reading into structured product building
  • Use a bounded development universe rather than only runtime or evidence entry points
How the three paths relate

LLM-side control, User-side evidence, and Development Platform can be adopted separately or together

Some teams start with the User-side because retained behavior records are the immediate need. Other teams start with the LLM-side because provider-facing execution is already becoming a production issue. Another group starts with the Development Platform because the immediate need is a dedicated workspace for structured building and implementation.

When used together, the LLM-side governs runtime behavior on the model path, the User-side preserves evidence first and extends into formal review later, and the Development Platform provides the bounded OC working environment for building and operating the product itself.

One side governs runtime behavior. One side preserves behavior evidence. One side provides the bounded development workspace.
Read the right implementation path

Go deeper only after choosing the correct side

LLM-side path

Choose this side when you need runtime control, governed execution, provider routing, runtime visibility, approval, and deeper operational control across provider-facing behavior. Runtime Integration is the concrete LLM-side start page after you choose this side.

Start LLM-side Open Core Runtime API Read Runtime Docs

User-side path

Choose this side when you need behavior evidence first, then later review built on retained records. Audit Workspace belongs to this review path, but access to it comes through Responsibility rather than standalone purchase. Start with the Evidence API start page when the goal is to get trial access, make the first write, and verify the first read.

Open Evidence API Start Open Evidence API Read Evidence Manual Open Audit Workspace

Development Platform path

Choose this side when your next step is opening the bounded OC build environment, using OC Workspace as the main working surface, and continuing structured implementation work inside CLARIXO.

Open OC Platform Open OC Workspace
Choose the path that matches your need

Start with the correct side, then expand when your system is ready

CLARIXO separates LLM-side runtime control, User-side evidence and later review, and Development Platform implementation work so teams can enter through the side that matches their immediate operational need.

Explore LLM-side Explore User-side Explore Development Platform