Skills
skills/fabioc-aloha/lithium/AIRS Integration

AIRS Integration

SKILL.md

Skill: AIRS Integration

Practical application of AIRS psychometric assessment for AI readiness, reliance calibration, and adoption optimization.

Metadata

Field Value
Skill ID airs-integration
Version 1.0.0
Category Research Tools
Difficulty Advanced
Prerequisites airs-appropriate-reliance
Related Skills cognitive-load-management, frustration-recognition

Overview

This skill bridges theoretical AIRS knowledge with practical application workflows. It enables Alex to conduct readiness assessments, calibrate reliance per-session, evaluate organizational deployment, and self-monitor for over/under-reliance patterns.

Core Value Proposition

AIRS (AI Readiness Scale) extends UTAUT2 to predict AI adoption with superior explanatory power. The key finding (Correa, 2025): Price Value (PV) β=.505 is the strongest predictor of Behavioral Intention — twice as powerful as any other construct.

Self-Assessment Link

For users seeking their own AIRS assessment:

Take the validated AIRS-16 at airs.correax.com/assessment

Module A: Project AI Readiness Assessment

Evaluate a user's project for AI integration readiness before investing development resources.

Readiness Dimensions

Dimension AIRS Construct Key Questions
Value Clarity Price Value (PV) Is the ROI clearly articulated? Can we quantify cost savings?
Technical Fit Effort Expectancy (EE) Will AI reduce complexity or add it? Integration friction?
Use Case Validity Performance Expectancy (PE) Will AI actually improve outcomes for this task?
User Readiness Hedonic Motivation (HM) Will users find the AI features engaging or threatening?
Social Dynamics Social Influence (SI) Do stakeholders champion or resist this AI integration?

Readiness Score Calculation

Project_Readiness = (PV_score × 2.0) + (EE_score × 1.5) + (PE_score × 1.2) + (HM_score × 0.8) + (SI_score × 0.5)
Max = 30 points

Weighting rationale: PV doubled because β=.505 is 2× stronger than other predictors.

Scoring Thresholds

Score Readiness Level Recommendation
24-30 High Readiness Proceed with AI integration
18-23 Moderate Readiness Address gaps before proceeding
12-17 Low Readiness Significant preparation needed
<12 Not Ready Fundamental issues — pause and reassess

Assessment Protocol

When user asks: "Is my project ready for AI?" / "Should we add AI to this?" / "AI readiness check"

Execute:

  1. Value Discovery (PV Focus)

    • "What cost savings or efficiency gains do you expect?"
    • "How would you measure ROI for this AI feature?"
    • Score 1-5 based on clarity of value proposition

  2. Effort Assessment (EE Focus)

    • "How complex is the integration with existing systems?"
    • "What's the learning curve for users?"
    • Score 1-5 based on friction level

  3. Performance Validation (PE Focus)

    • "What specific outcomes will improve with AI?"
    • "Do you have baseline metrics to compare against?"
    • Score 1-5 based on outcome clarity

  4. Engagement Potential (HM Focus)

    • "Will users find this AI feature enjoyable or intimidating?"
    • "Is there existing enthusiasm for AI in this context?"
    • Score 1-5 based on user receptivity

  5. Stakeholder Map (SI Focus)

    • "Who are the AI champions in this project?"
    • "What resistance have you encountered?"
    • Score 1-5 based on social support

  6. Calculate & Report

    • Apply weighted formula
    • Identify weakest dimension
    • Provide targeted recommendations

Module B: Session Reliance Calibration

Optimize user reliance on Alex during active sessions using real-time calibration.

The Calibration Problem

Pattern Description Risk
Over-reliance Accepting Alex outputs without verification Errors propagate silently
Under-reliance Ignoring Alex suggestions, doing everything manually Efficiency loss, fatigue
Appropriate Reliance Verifying when uncertain, trusting when confident Optimal outcomes

CAIR/CSR Framework (Schemmer et al., 2023)

User Accepts User Rejects
AI Correct ✅ CAIR (Correct AI-Reliance) ❌ Under-reliance
AI Incorrect ❌ Over-reliance ✅ CSR (Correct Self-Reliance)

Session Calibration Signals

Over-reliance indicators:

  • Accepting all suggestions without modifications
  • Not verifying AI-generated code before running
  • Asking Alex to make decisions user should own
  • "Just do it" without reviewing approach

Under-reliance indicators:

  • Ignoring Alex suggestions repeatedly
  • Manually typing code Alex already generated
  • Rejecting help before evaluating it
  • "I'll do it myself" after requesting help

Calibration Interventions

For over-reliance (High TR, Low AR):

"I notice you're trusting my outputs quickly. For this [critical/complex] task, would you like to review together? I want to make sure we catch any issues early."

For under-reliance (Low TR, Any AR):

"I see you're preferring to work manually. That's totally valid — but I could help with [specific subtask] to save time. Would you like to try a hybrid approach?"

Session Start Calibration Check

At session start (or when /calibrate invoked):

  1. "On a scale of 1-5, how much are you planning to rely on AI for today's work?"
  2. "What's your verification strategy for AI-generated output today?"
  3. Adjust interaction style based on answers:
    • High reliance intent → Add more "let me explain my reasoning" moments
    • Low reliance intent → Be more assertive about high-confidence suggestions

Module C: Enterprise Deployment Readiness

Assess organizational readiness for AI tool deployment at scale.

The Price Value Imperative

Critical insight: PV (β=.505) suggests organizations must lead with ROI clarity. "AI is cool" fails. "AI will save 40% on X" succeeds.

Enterprise Readiness Framework

Domain Assessment Focus AIRS Link
Business Case Clear, quantified value proposition PV (primary driver)
Technical Infrastructure Integration readiness, data quality EE
Change Management Training programs, champion network FC, SI
User Experience Ease of adoption, enjoyment potential EE, HM
Governance Policies, ethics frameworks, oversight FC

Typology-Based Deployment Strategy

From AIRS research (Correa, 2025):

User Type AIRS Score Deployment Role Strategy
AI Enthusiasts >30 Early adopters, champions Recruit as internal advocates, beta testers
Moderate Users 21-30 Mainstream adoption Lead with value proof, reduce friction
AI Skeptics ≤20 Late adoption Peer influence, gradual exposure, choice

Enterprise Assessment Protocol

When user asks: "Is my org ready for AI?" / "AI deployment assessment" / "Enterprise readiness"

Execute:

  1. Business Case Review (PV Focus)

    • "Do you have a quantified ROI for this AI deployment?"
    • "What's the cost of NOT implementing AI?"
    • Red flag: No clear numbers = Not ready

  2. Technical Readiness (EE Focus)

    • "How will AI integrate with existing workflows?"
    • "What data quality issues exist?"
    • Red flag: Major integration gaps = High risk

  3. Change Readiness (SI + FC Focus)

    • "Who are your AI champions at each level?"
    • "What training/support is planned?"
    • Red flag: No champion network = Adoption failure likely

  4. User Segmentation (All Constructs)

  5. Governance Check (FC Focus)

    • "What AI policies are in place?"
    • "Who owns AI ethics decisions?"
    • Red flag: No governance = Liability risk

Deployment Recommendation Matrix

Business Case Technical Ready Change Ready Recommendation
✅ ✅ ✅ Full deployment
✅ ✅ ❌ Pilot with champions
✅ ❌ ✅ Technical sprint first
❌ Any Any STOP — build business case

Module D: Alex Self-Monitoring

Enable Alex to detect and correct reliance imbalances in real-time.

Self-Monitoring Triggers

Signal Detection Response
User accepts 5+ suggestions without edit Counter tracking Insert calibration moment
User rejects 3+ suggestions in a row Rejection tracking Offer meta-conversation
Complex task, no verification request Task complexity scan Proactively suggest review
User asks "just do it" on critical task Phrase detection Pause and confirm scope

Calibration Moments

Periodic check-in (every 15-20 interactions in active session):

"Quick calibration check: Are you finding my suggestions on-target today? Anything I'm missing or over-explaining?"

Post-error recovery:

"That didn't work as expected. Let's pause — what should I have caught earlier? This helps me calibrate better."

Session wrap-up:

"Before we wrap: Any patterns you noticed in my suggestions? Times I was over-confident or too cautious?"

Reliance Metrics (Per-Session)

Metric Formula Target
Acceptance Rate Accepted / Total Suggestions 60-80% (not too high, not too low)
Modification Rate Modified / Accepted 20-40% (healthy verification)
Rejection Rate Rejected / Total 10-30% (healthy skepticism)

Interpretation:

  • Acceptance >90% → Over-reliance concern
  • Rejection >50% → Under-reliance or poor fit
  • Modification 20-40% → Healthy calibration

Self-Improvement Loop

When calibration issues detected:

  1. Log pattern to episodic memory
  2. Identify root cause (domain mismatch, explanation style, confidence miscommunication)
  3. Adjust behavior in current session
  4. If pattern persists, surface in meditation for architectural review

Integration Workflows

Complete AIRS Assessment Flow

%%{init: {'theme': 'base', 'themeVariables': {
  'primaryColor': '#cce5ff',
  'primaryTextColor': '#333',
  'primaryBorderColor': '#57606a',
  'lineColor': '#57606a',
  'secondaryColor': '#e6d5f2',
  'tertiaryColor': '#c2f0d8',
  'background': '#ffffff',
  'mainBkg': '#cce5ff',
  'secondBkg': '#e6d5f2',
  'tertiaryBkg': '#c2f0d8',
  'textColor': '#333',
  'border1Color': '#57606a',
  'border2Color': '#57606a',
  'arrowheadColor': '#57606a',
  'fontFamily': 'ui-sans-serif, system-ui, sans-serif',
  'fontSize': '14px',
  'nodeBorder': '1.5px',
  'clusterBkg': '#f6f8fa',
  'clusterBorder': '#d0d7de',
  'edgeLabelBackground': '#ffffff'
}}}%%
flowchart TD
    A[User Request] --> B{Assessment Type?}
    B -->|Project| C[Module A: Project Readiness]
    B -->|Personal| D[airs.correax.com/assessment]
    B -->|Enterprise| E[Module C: Enterprise Readiness]
    B -->|Session| F[Module B: Calibration Check]
    
    C --> G[Weighted Score]
    D --> H[AIRS-16 Results + History]
    E --> I[Deployment Matrix]
    F --> J[Reliance Adjustment]
    
    G --> K[Recommendations]
    H --> K
    I --> K
    J --> K

Key Links

Resource URL Purpose
Personal Assessment airs.correax.com/assessment Take AIRS-16
Assessment History airs.correax.com/history Track changes over time
Org Deployment airs.correax.com/org/register Enterprise surveys

Quick Reference

AIRS-16 Constructs

Construct Key Finding Application
Price Value (PV) β=.505 (strongest) Lead with ROI — always
Performance Expectancy (PE) Significant Demonstrate outcomes
Effort Expectancy (EE) Significant Reduce friction
Hedonic Motivation (HM) Significant Make it enjoyable
Social Influence (SI) Significant Leverage champions
Habit (HA) Significant Build into workflow
Facilitating Conditions (FC) Significant Ensure support
Trust (TR) Marginal (p>.05) Trust alone ≠ adoption

The Trust Paradox

Trust (TR) was marginal in AIRS-16 validation. This suggests:

  • Trust level alone doesn't predict adoption
  • Calibration (knowing when to trust) matters more
  • See airs-appropriate-reliance skill for AR extension hypothesis

Activation Patterns

Trigger Response
"AI readiness", "ready for AI", "should we add AI" Module A: Project Assessment
"calibrate", "reliance check", "am I over-relying" Module B: Session Calibration
"enterprise AI", "org deployment", "scale AI" Module C: Enterprise Assessment
/calibrate command Session calibration check
High acceptance rate detected Self-monitoring intervention
Complex task without verification Proactive calibration moment

References

  • Correa, F. (2025). AIRS-16 Validation Study (N=523, CFI=.975)
  • Schemmer, M., et al. (2023). CAIR/CSR Framework for Appropriate Reliance
  • Passi, S., Dhanorkar, S., & Vorvoreanu, M. (2024). AETHER: Appropriate Reliance Synthesis
  • Venkatesh, V., et al. (2012). UTAUT2 Model

Skill created: 2026-02-10 | Category: Research Tools | Status: Active

Synapses

  • [.github/instructions/empirical-validation.instructions.md] (Critical, Validates, Bidirectional) - "AIRS is empirically validated scale"
  • [.github/instructions/bootstrap-learning.instructions.md] (High, Measures, Forward) - "Calibrate learning effectiveness"
  • [.github/skills/airs-appropriate-reliance/SKILL.md] (High, Extends, Forward) - "Extended AR framework"
Weekly Installs
0
Repository
fabioc-aloha/lithium
First Seen
Jan 1, 1970
Security Audits
Gen Agent Trust HubPass
SocketPass
SnykPass