Component Library — Review Draft

Claire Reusable Component Map

38 pre-built components organized into 5 tiers. Each component is reusable across multiple use cases. Team: please review and flag anything missing, incorrect, or misplaced.

Tier 1: Connector Tools
Tier 2: Domain Skills
Tier 3: Specialized Agents
Tier 4: Reusable Flows
Tier 5: Ontology
|
Built
Partial
Planned
1
Connector Tools
Data plumbing — pull data from external systems. Every use case needs at least one.
HTTP Tool
🏥
Veeva Vault
Audit trail, documents, workflows, record lifecycle
AuditMigrationQuality
Partial
HTTP Tool
🎫
ServiceNow
Change requests, incidents, access requests, approvals
AuditCAPAChange Ctrl
Planned
HTTP Tool
📋
JIRA
Issues, workflows, sprint tracking, custom fields
AuditCAPAChange Ctrl
Planned
HTTP Tool
🏭
SAP QM
Quality notifications, inspection lots, batch data
ManufacturingQuality
Planned
HTTP Tool
🔍
TrackWise
CAPAs, deviations, complaints, audit findings
CAPADeviationComplaints
Planned
HTTP Tool
📁
MasterControl
Document control, training records, change orders
QualityTraining
Planned
Built-in
📄
PDF Extract
Extract text from PDF documents, organized by page
MigrationBatch ReviewSubmissions
Built
Built-in
🕸️
Graph Query
Cypher queries against Neo4j Knowledge Graph
Everything
Built
🔍 Team Review Needed

For each connector: What API version? What endpoints? What auth method (OAuth, session, API key)? Any existing integration code or Postman collections we can reuse? Are there systems we're missing?

2
Domain Skills (Reusable Prompt Templates)
Domain expertise packaged as prompt templates. Each skill can be attached to any agent. The quality of these prompts determines the quality of Claire's output.
Skill
📖
SOP Compliance Check
Compare an activity against its governing SOP. Returns: COMPLIANT, DEVIATION, or CRITICAL with SOP section reference.
AuditCAPADeviationChange Ctrl
Planned
Skill
🔗
Ticket Correlation
Match activities to change/access tickets by record ID, time window, and user. Flag unmatched activities.
AuditCAPAChange Ctrl
Planned
Skill
🔮
Pattern Detection
Find anomalous patterns: off-hours activity, bulk operations, privilege escalation, repeated failures.
AuditSafetyPharmacovigilance
Planned
Skill
⚖️
Risk Scoring
Assess risk using probability × severity matrix. Outputs: LOW, MEDIUM, HIGH, CRITICAL with justification.
CAPADeviationChange CtrlSafety
Planned
Skill
📝
Report Generator
Format findings into a formal compliance report with executive summary, findings, and recommendations.
Everything
Planned
Skill
🔬
Root Cause Analysis
Perform 5-Why or Fishbone analysis from incident data. Identify contributing factors and root cause.
CAPADeviation
Planned
Skill
🏷️
Regulatory Classification
Classify content per FDA/EMA regulatory categories, submission types, or document taxonomies.
SubmissionsIntelligenceMigration
Planned
Skill
📊
Gap Analysis
Compare current state vs requirements. Identify gaps in SOPs, training, or processes.
AuditQualityTraining
Planned
Skill
🔐
Authorization Check
Verify user had the required role/permission to perform an action. Cross-reference with access requests.
AuditChange CtrlAccess Review
Planned
Skill
📋
Evidence Summarizer
Summarize supporting evidence for findings — condense logs, documents, and data into concise narratives.
CAPAAuditDeviation
Planned
Skill
⏱️
Timeline Reconstructor
Build chronological narrative from disparate events. Create an "event timeline" for investigations.
InvestigationDeviationCAPA
Planned
Skill
Effectiveness Check
Evaluate if a corrective action was effective. Compare before/after metrics and recurrence rate.
CAPA follow-up
Planned
🔍 Team Review Needed

For each skill: Is the description accurate? Are we using the right methodology (e.g., 5-Why vs Fishbone for RCA)? What risk matrix do clients use? Can you share a sample SOP, audit report, or CAPA form so we can write accurate prompt templates?

3
Specialized Agents
Each agent has a purpose, uses specific tools, and applies specific skills. Agents are composed from Tier 1 + Tier 2 components.
Agent
🔍
Audit Trail Reviewer
Reviews system audit logs for compliance. Detects gaps, unauthorized changes, SOP deviations.
Tools: Veeva Connector, Graph Query
Skills: SOP Check, Pattern Detection, Ticket Correlation, Report Generator
iQuality
Planned
Agent
📄
Document Classifier
Classifies documents by type and regulatory category. Maps metadata fields for migration.
Tools: PDF Extract
Skills: Regulatory Classification
MAiGRATEQuality
Partial
Agent
🔧
CAPA Investigator
Investigates root cause, assesses risk, recommends corrective actions, drafts effectiveness plan.
Tools: SNOW/JIRA/TrackWise Connector
Skills: Root Cause Analysis, Risk Scoring, Evidence Summarizer, Timeline Reconstructor
iQuality
Planned
Agent
⚠️
Deviation Analyst
Analyzes deviations — timeline, root cause, risk assessment, impact on batches/products.
Tools: Veeva/SAP Connector
Skills: Risk Scoring, Gap Analysis, Timeline Reconstructor
iQualityManufacturing
Planned
Agent
🔄
Change Control Reviewer
Reviews change requests for completeness, SOP compliance, risk, and authorization.
Tools: SNOW/JIRA Connector
Skills: SOP Check, Authorization Check, Risk Scoring
iQualityChange Mgmt
Planned
Agent
🎓
Training Compliance Monitor
Tracks training completion, identifies gaps vs role requirements, flags overdue training.
Tools: MasterControl/SharePoint Connector
Skills: Gap Analysis, Report Generator
QualityTraining
Planned
Agent
📰
Regulatory Intelligence
Monitors new guidelines, classifies regulatory updates, assesses impact on existing SOPs.
Tools: Web/RSS Tool
Skills: Regulatory Classification, Gap Analysis
Regulatory
Planned
Agent
🏭
Batch Record Reviewer
Reviews batch manufacturing records for GMP compliance — checks procedures, signatures, deviations.
Tools: SAP/Veeva Connector, PDF Extract
Skills: SOP Check, Pattern Detection
Manufacturing
Planned
Agent
📞
Complaint Handler
Triages product complaints, investigates root cause, assesses risk, generates response.
Tools: TrackWise/SNOW Connector
Skills: Risk Scoring, Root Cause Analysis, Timeline Reconstructor
QualityPost-Market
Planned
Agent
🏢
Supplier Qualifier
Assesses supplier qualification status — checks documentation, audit history, risk profile.
Tools: SharePoint/Custom DB
Skills: Risk Scoring, Gap Analysis, Evidence Summarizer
ProcurementQuality
Planned
🔍 Team Review Needed

For each agent: Is the purpose description accurate? Are the right tools/skills attached? Are we missing any agents that clients frequently need? Should any agents be split or merged?

How Components Compose Together
🔧 Tools
Veeva, SNOW, PDF
+
🎯 Skills
SOP Check, Risk, RCA
🤖 Agent
CAPA Investigator
⚡ Flow
CAPA Investigation
🏪 Store
One-click install
Each tier builds on the previous. Tools + Skills → Agents → Flows → Published to Store for any realm to install.
4
Reusable Flows (End-to-End Workflows)
Complete workflows that chain agents together. Install a flow and get everything it depends on (agents, tools, skills).
Flow
🔍
Audit Trail Review
Pull logs → Parse → Query Graph → Deterministic Rules → AI Review → Report
Agents: Audit Trail Reviewer
Planned
Flow
🔧
CAPA Investigation
Receive deviation → Investigate → Root Cause → Corrective Action → Effectiveness Plan
Agents: CAPA Investigator
Planned
Flow
🔄
Change Control Assessment
Receive request → Check completeness → Impact analysis → SOP review → Decision
Agents: Change Control Reviewer
Planned
Flow
📄
Document Migration
Upload PDF → Classify → Extract metadata → Map fields → Upload to Vault
Agents: Document Classifier
Partial
Flow
⚠️
Deviation Investigation
Receive deviation → Timeline → Root cause → Risk assessment → Report
Agents: Deviation Analyst
Planned
Flow
🎓
Training Gap Analysis
Pull training records → Compare vs requirements → Identify gaps → Report
Agents: Training Compliance Monitor
Planned
Flow
📰
Regulatory Impact Assessment
New guideline → Classify → Find affected SOPs → Gap analysis → Report
Agents: Regulatory Intelligence
Planned
Flow
📞
Complaint Processing
Receive complaint → Triage → Investigate → Risk assess → Report
Agents: Complaint Handler
Planned
🔍 Team Review Needed

For each flow: Is the step sequence correct? Are we missing steps? Which flows should we build first based on client demand? Are there flows we haven't thought of?

5
Universal Life Sciences Ontology
Knowledge Graph schema that connects everything. Agents traverse these relationships instead of making API calls.
Entity Relationships
Core Entities
System ──GOVERNED_BY──→ SOP
SOP ──DEFINES──→ Procedure
Procedure ──REQUIRES_ROLE──→ Role
User ──HAS_ROLE──→ Role
User ──AUTHORIZED_ON──→ System
Ticket ──RAISED_FOR──→ System
Ticket ──APPROVED_BY──→ User
Extended Entities
Product ──HAS_BATCH──→ Batch
Deviation ──AFFECTS──→ Batch
CAPA ──ADDRESSES──→ Deviation
Complaint ──ABOUT──→ Product
Supplier ──SUPPLIES──→ Product
Training ──REQUIRED_BY──→ Role
Training ──COMPLETED_BY──→ User
🔍 Team Review Needed

Ontology review: Are the entity types correct? Are we using the right terminology? (CAPA vs Corrective Action?) Are relationships accurate? Any entities or relationships missing? Does CG have an existing data model we should align with?


Claire Agentic AI Studio — Component Library Review Draft
Please share feedback directly on this document or tag @phanincg