Changes for page Workflows

Last modified by Robert Schaub on 2025/12/24 20:34

From 7.1 to 6.1 From 7.4 to 7.3

From version 7.3

edited by Robert Schaub
on 2025/12/16 20:28

Change comment: Renamed back-links.

To version 7.1

edited by Robert Schaub
on 2025/12/15 16:56

Change comment: Imported from XAR

Raw
Rendered

Summary

Page properties (1 modified, 0 added, 0 removed)

Details

Page properties

Content

@@ -21,7 +21,6 @@
  **Actor**: Contributor or AKEL
  **Actions**:
--
  * Submit claim text
  * Provide initial sources (optional for human contributors, mandatory for AKEL)
  * System assigns initial AuthorType (Human or AI)
@@ -31,7 +31,6 @@
  === Step 2: AKEL Processing ===
  **Automated Steps**:
--
 . Claim extraction and normalization
 . Classification (domain, type, evaluability)
 . Risk tier assignment (A/B/C suggested)
@@ -45,7 +45,6 @@
  === Step 3: Quality Gate Checkpoint ===
  **Gates Evaluated**:
--
  * Source quality
  * Contradiction search completion
  * Uncertainty quantification
@@ -52,7 +52,6 @@
  * Structural integrity
  **Outcomes**:
--
  * **All gates pass** → Proceed to Mode 2 publication (if Tier B or C)
  * **Any gate fails** → Mode 1 (Draft), flag for human review
  * **Tier A** → Mode 2 with warnings + auto-escalate to expert queue
@@ -60,19 +60,16 @@
  === Step 4: Publication (Risk-Tier Dependent) ===
  **Tier C (Low Risk)**:
--
  * **Direct to Mode 2**: AI-generated, public, clearly labeled
  * User can request human review
  * Sampling audit applies
  **Tier B (Medium Risk)**:
--
  * **Direct to Mode 2**: AI-generated, public, clearly labeled
  * Higher audit sampling rate
  * High-engagement content may auto-escalate
  **Tier A (High Risk)**:
--
  * **Mode 2 with warnings**: AI-generated, public, prominent disclaimers
  * **Auto-escalated** to expert review queue
  * User warnings displayed
@@ -81,7 +81,6 @@
  === Step 5: Human Review (Optional for B/C, Escalated for A) ===
  **Triggers**:
--
  * User requests review
  * Audit flags issues
  * High engagement (Tier B)
@@ -88,7 +88,6 @@
  * Automatic (Tier A)
  **Process**:
--
 . Reviewer/Expert examines claim
 . Validates quality gates
 . Checks contradiction search results
@@ -96,7 +96,6 @@
 . Decision: Approve, Request Changes, or Reject
  **Outcomes**:
--
  * **Approved** → Mode 3 (Human-Reviewed)
  * **Changes Requested** → Back to contributor or AKEL for revision
  * **Rejected** → Rejected status with reasoning
@@ -108,13 +108,11 @@
  === Step 1: Scenario Generation ===
  **Automated (AKEL)**:
--
  * Generate scenarios for claim
  * Define boundaries, assumptions, context
  * Identify evaluation methods
  **Manual (Expert/Reviewer)**:
--
  * Create custom scenarios
  * Refine AKEL-generated scenarios
  * Add domain-specific nuances
@@ -122,7 +122,6 @@
  === Step 2: Scenario Validation ===
  **Quality Checks**:
--
  * Completeness (definitions, boundaries, assumptions clear)
  * Relevance to claim
  * Evaluability
@@ -129,7 +129,6 @@
  * No circular logic
  **Risk Tier Assignment**:
--
  * Inherits from parent claim
  * Can be overridden by expert if scenario increases/decreases risk
@@ -136,12 +136,10 @@
  === Step 3: Scenario Publication ===
  **Mode 2 (AI-Generated)**:
--
  * Tier B/C scenarios can publish immediately
  * Subject to sampling audits
  **Mode 1 (Draft)**:
--
  * Tier A scenarios default to draft
  * Require expert validation for Mode 2 or Mode 3
@@ -152,7 +152,6 @@
  === Step 1: Evidence Search & Retrieval ===
  **AKEL Actions**:
--
  * Search academic databases, reputable media
  * **Mandatory contradiction search** (counter-evidence, reservations)
  * Extract metadata (author, date, publication, methodology)
@@ -159,7 +159,6 @@
  * Assess source reliability
  **Quality Requirements**:
--
  * Primary sources preferred
  * Diverse perspectives included
  * Echo chambers flagged
@@ -168,7 +168,6 @@
  === Step 2: Evidence Summarization ===
  **AKEL Generates**:
--
  * Summary of evidence
  * Relevance assessment
  * Reliability score
@@ -180,7 +180,6 @@
  === Step 3: Evidence Review ===
  **Reviewer/Expert Validates**:
--
  * Accuracy of summaries
  * Appropriateness of sources
  * Completeness of contradiction search
@@ -187,7 +187,6 @@
  * Reliability assessments
  **Outcomes**:
--
  * **Mode 2**: Evidence summaries published as AI-generated
  * **Mode 3**: After human validation
  * **Mode 1**: Failed quality checks or pending expert review
@@ -199,7 +199,6 @@
  === Step 1: Verdict Computation ===
  **AKEL Computes**:
--
  * Verdict across scenarios
  * Confidence scores
  * Uncertainty quantification
@@ -207,7 +207,6 @@
  * Limitations
  **Inputs**:
--
  * Claim text
  * Scenario definitions
  * Evidence assessments
@@ -216,13 +216,11 @@
  === Step 2: Verdict Validation ===
  **Quality Gates**:
--
  * All four gates apply (source, contradiction, uncertainty, structure)
  * Reasoning chain must be traceable
  * Assumptions must be explicit
  **Risk Tier Check**:
--
  * Tier A: Always requires expert validation for Mode 3
  * Tier B: Mode 2 allowed, audit sampling
  * Tier C: Mode 2 default
@@ -230,7 +230,6 @@
  === Step 3: Verdict Publication ===
  **Mode 2 (AI-Generated Verdict)**:
--
  * Clear labeling with confidence scores
  * Uncertainty disclosure
  * Links to reasoning trail
@@ -237,7 +237,6 @@
  * User can request expert review
  **Mode 3 (Expert-Validated Verdict)**:
--
  * Human reviewer/expert stamp
  * Additional commentary (optional)
  * Highest trust level
@@ -249,7 +249,6 @@
  === Step 1: Audit Sampling Selection ===
  **Stratified Sampling**:
--
  * Risk tier priority (A > B > C)
  * Low confidence scores
  * High traffic content
@@ -257,7 +257,6 @@
  * User flags
  **Sampling Rates** (Recommendations):
--
  * Tier A: 30-50%
  * Tier B: 10-20%
  * Tier C: 5-10%
@@ -265,7 +265,6 @@
  === Step 2: Audit Execution ===
  **Auditor Actions**:
--
 . Review sampled AI-generated content
 . Validate quality gates were properly applied
 . Check contradiction search completeness
@@ -273,7 +273,6 @@
 . Identify errors or hallucinations
  **Audit Outcome**:
--
  * **Pass**: Content remains in Mode 2, logged as validated
  * **Fail**: Content flagged for review, system improvement triggered
@@ -280,7 +280,6 @@
  === Step 3: Feedback Loop ===
  **System Improvements**:
--
  * Failed audits analyzed for patterns
  * AKEL parameters adjusted
  * Quality gates refined
@@ -287,7 +287,6 @@
  * Risk tier assignments recalibrated
  **Transparency**:
--
  * Audit statistics published periodically
  * Patterns shared with community
  * System improvements documented
@@ -299,7 +299,6 @@
  === Mode 1 → Mode 2 ===
  **Requirements**:
--
  * All quality gates pass
  * Risk tier B or C (or A with warnings)
  * Contradiction search completed
@@ -309,7 +309,6 @@
  === Mode 2 → Mode 3 ===
  **Requirements**:
--
  * Human reviewer/expert validation
  * Quality standards confirmed
  * For Tier A: Expert approval required
@@ -320,13 +320,11 @@
  === Mode 3 → Mode 1 (Demotion) ===
  **Rare - Only if**:
--
  * New evidence contradicts verdict
  * Error discovered in reasoning
  * Source retraction
  **Process**:
--
 . Content flagged for re-evaluation
 . Moved to draft (Mode 1)
 . Re-processed through workflow
@@ -339,12 +339,10 @@
  === On Mode 1 (Draft) Content ===
  **Contributors**:
--
  * Edit their own drafts
  * Submit for review
  **Reviewers/Experts**:
--
  * View and comment
  * Request changes
  * Approve for Mode 2 or Mode 3
@@ -352,7 +352,6 @@
  === On Mode 2 (AI-Generated) Content ===
  **All Users**:
--
  * Read and use content
  * Request human review
  * Flag for expert attention
@@ -359,7 +359,6 @@
  * Provide feedback
  **Reviewers/Experts**:
--
  * Validate for Mode 3 transition
  * Edit and refine
  * Adjust risk tier if needed
@@ -367,12 +367,10 @@
  === On Mode 3 (Human-Reviewed) Content ===
  **All Users**:
--
  * Read with highest confidence
  * Still can flag if new evidence emerges
  **Reviewers/Experts**:
--
  * Update if needed
  * Trigger re-evaluation if new evidence
@@ -382,11 +382,11 @@
  === Claim and Scenario Lifecycle (Overview) ===
--{{include reference="FactHarbor.Archive.FactHarbor V0\.9\.23 Lost Data.Organisation.Diagrams.Claim and Scenario Lifecycle (Overview).WebHome"/}}
++{{include reference="Test.FactHarborV09.Organisation.Diagrams.Claim and Scenario Lifecycle (Overview).WebHome"/}}
  === Claim and Scenario Workflow ===
--{{include reference="FactHarbor.Archive.FactHarbor V0\.9\.23 Lost Data.Specification.Diagrams.Claim and Scenario Workflow.WebHome"/}}
++{{include reference="Test.FactHarborV09.Specification.Diagrams.Claim and Scenario Workflow.WebHome"/}}
  === Evidence and Verdict Workflow ===
@@ -408,3 +408,4 @@
  * [[Automation>>FactHarbor.Specification.Automation.WebHome]]
  * [[Requirements (Roles)>>FactHarbor.Specification.Requirements.WebHome]]
  * [[Governance>>FactHarbor.Organisation.Governance]]
++

Changes for page Workflows

Summary

Details

Applications

Navigation

Need help?