Skip to content

FEATURE_MATRIX.md

Latest commit

 

History

History
115 lines (84 loc) · 4.63 KB

FEATURE_MATRIX.md

File metadata and controls

115 lines (84 loc) · 4.63 KB

IOA Core Feature Matrix

Version: 2.5.1
Last Updated: 2025-10-10

This document lists verified features in the current public IOA Core release candidate.

Core Features

✅ Audit & Governance

Feature Status Description
Audit Logging ✅ Available Immutable audit chains with cryptographic verification
Audit Chain Verification ✅ Available Verify audit chain integrity and sequence
Data Redaction ✅ Available Automatic redaction of sensitive data in logs
System Laws Framework ✅ Available Seven governing principles for AI orchestration
Ethics Frameworks ✅ Available Reference integration with established ethics assessment tools (Aletheia v2.0 attribution; no derivatives)

✅ Memory System

Feature Status Description
Hot Storage ✅ Available In-memory storage for active data
Cold Storage ✅ Available Persistent storage for archived data
Multiple Backends ✅ Available SQLite, S3, Local JSONL
Encryption (AES-GCM) ✅ Available At-rest encryption for memory
4D Tiering ⚠️ Preview Educational preview of tiered memory management

✅ LLM Provider Support

Provider Status Notes
OpenAI ✅ Available GPT-4, GPT-3.5 support
Anthropic ✅ Available Claude 3.x support
Google Gemini ✅ Available Gemini 1.5 support
DeepSeek ✅ Available DeepSeek Chat support
XAI ✅ Available Grok Beta support
Ollama ✅ Available Local model support with turbo mode

Ollama Turbo Mode: Optimized preset for faster local inference (20-40% latency improvement on capable hardware, results vary).

✅ CLI Interface

Command Status Description
ioa --help ✅ Available Main CLI help
ioa smoketest-live ✅ Available Live provider smoke testing
Provider management ✅ Available Configure and test LLM providers

✅ Ethics Cartridge (Preview)

Feature Status Description
Ethics Precheck ✅ Available Runtime ethics validation with confidence scoring
PII Detection ✅ Available Basic detection of personally identifiable information
Deception Detection ✅ Available Identification of potential manipulation attempts
Harmful Content Detection ✅ Available Basic screening for inappropriate content
Fairness Violation Detection ✅ Available Detection of potential bias or discrimination
Policy Validation ✅ Available Validate ethics policy configurations

Note: Ethics cartridge (Aletheia-inspired) - IOA original implementation with neutral criteria names. Not a derivative of Aletheia Framework v2.0.

Separate Commercial Extensions

The following capabilities are not part of the public IOA Core repository and should not be implied by OSS-facing docs:

  • Workflow Engine (YAML DSL) - Visual workflow builder
  • PKI Agent Onboarding - Advanced agent management
  • Multi-region Replication - Geo-distributed storage
  • Advanced Analytics - Detailed usage and performance analytics
  • Compliance Frameworks - Full HIPAA, SOX, GDPR tooling
  • Visual Builder - Drag-and-drop workflow designer

Feature Verification

All features listed as "✅ Available" have been verified with proof tests:

  • tests/feature_sync/test_audit_chain.py - Audit logging
  • tests/feature_sync/test_memory_fabric.py - Memory system
  • tests/feature_sync/test_encryption.py - AES-GCM encryption
  • tests/feature_sync/test_providers.py - LLM providers
  • tests/perf/test_ollama_turbo.py - Ollama turbo mode

Run proof tests:

python3 tests/feature_sync/test_*.py
python3 tests/perf/test_ollama_turbo.py

Core vs Commercial Extensions

Category Public IOA Core Commercial Extensions
Core Engine ✅ Full ✅ Full
LLM Providers ✅ 6 providers ✅ 6+ providers
Governance ✅ Basic ✅ Advanced
Memory System ✅ Hot/Cold ✅ Multi-tier + Advanced
Workflow Engine ❌ Future ✅ Visual Builder
Compliance ✅ Educational ✅ Production-Ready

Note: Public IOA Core currently focuses on audit, evidence, memory, and multi-provider orchestration primitives. Commercial offerings may add hosted operations, packaged compliance content, advanced analytics, or visual tooling.

For more information, see README.md.