CLARIXO
Audit Workspace Protocol

Define the role, boundary, and minimum integration contract for formal AI review work.

CLARIXO Audit Workspace is built on top of evidence records that already exist. It is used to open a formal case, define a bounded evidence scope, preserve reviewer context, and support structured delivery without mutating source evidence.

Open Audit Workspace Read Manual Open Evidence API Back to Docs
What It Is

A structured case-review workspace built above evidence records.

Audit Workspace helps teams move from “evidence exists” to “evidence can be grouped, reviewed, explained, and prepared for delivery.” It is designed for grouped review, reviewer context, formal note-taking, and export preparation around user-facing AI behavior that has already been recorded.

Role

Open and carry a formal review case

Use Audit Workspace to define case identity, preserve review continuity, and keep reviewer context separate from source evidence.

Boundary

Review layer, not runtime control

Audit Workspace does not replace Core Runtime API and does not replace Evidence API. It depends on evidence that already exists.

What It Is Not

Not a routing layer, not an enforcement layer, and not an automated verdict system.

Not Runtime Control
Audit Workspace does not own routing, fallback, execution selection, provider switching, or guard enforcement.
Not Evidence Ingestion
It should not be used as the place where evidence payload structure is invented for the first time. Evidence should already be stable before case review begins.
Not Liability Judgment
It supports structured examination and formal packaging. It does not issue legal conclusions or automated responsibility findings.
Where It Sits

CLARIXO now separates runtime control, behavior evidence, and formal review.

Core Runtime API
The LLM-side runtime family between applications and external AI providers. This is where routing, fallback, guard behavior, approval, and execution governance belong.
Evidence API
The user-side foundation layer for exporting user-facing AI behavior as traceable, provable, and auditable records.
Audit Workspace
The review workspace for grouped review, reviewer context, formal notes, and export preparation built on evidence records.
Audit Workspace depends on the evidence layer. The minimum principle is simple: record first, review second.
Prerequisites

Evidence must already be stable before a formal case is opened.

Project Context
The integrating team should already have a valid CLARIXO project context that can anchor case review.
Evidence Availability
User-facing AI behavior should already be written into CLARIXO through the evidence path and be queryable by project, application, session, or time range.
Review Target
The team should already know which event cluster, interaction sequence, or review window needs formal examination.
Minimum Integration Principle

Open a case only after the evidence path is already working.

Step 1
Export evidence through the evidence path and confirm that it can be read back under the expected project and application context.
Step 2
Open a case in Audit Workspace and define a narrow, bounded evidence scope for first review.
Step 3
Review the grouped evidence inside Case Detail, preserve reviewer context, and export a formal packet when needed.
Case Model

A case is a bounded review container, not a mutation of source evidence.

Case Identity
A case should carry a title and stable scope anchors such as project, application, optional session, and recorded-from / recorded-to boundaries.
Scope
The scope should remain narrow, stable, and explainable, especially for first-case validation and formal export quality.
Separation
Evidence records remain in the evidence layer. Case records, notes, and export logs remain in Audit Workspace.
Human Review Principle

Evidence supports review, but does not eliminate human judgment.

Audit Workspace is built for assisted review rather than automatic conclusion. Reviewers may summarize, annotate, and package evidence for downstream use, but they should not confuse structured evidence with automatic liability determination.

Minimum Successful Outcome

An integration is operational when the full review chain works end to end.

Open
A real case can be opened under a valid project context.
Read
Grouped evidence can be read inside Case Detail according to the scoped boundary.
Review
Reviewer summary and notes can be preserved without mutating underlying evidence.
Export
A formal packet can be generated while preserving case-to-evidence continuity.
Closing Statement

Audit Workspace is the structured review workspace between evidence collection and downstream use.

It is not the control layer and not the verdict layer. Its role is to help organizations open, bound, review, explain, and prepare grouped evidence for delivery when responsibility-sensitive workflows become part of the operating model.

Continue to Manual Open Audit Workspace Back to Docs