ShieldCortex
The security layer for AI agent memory. Like Cloudflare, but for everything your AI remembers. 6-layer defence pipeline, Iron Dome behavioural protection, credential leak detection, and universal memory bridge. One package. Full solution.
Free & open source — MIT licensed
Defence In Depth
Multiple layers of protection for your AI agent's memory system
Memory Firewall
Three modes — monitor, filter, and block — to control what gets stored in agent memory.
Trust Scoring
Assigns trust scores to memory entries based on source, content analysis, and historical patterns.
Audit Logger
Complete audit trail of every memory operation — reads, writes, deletes, and modifications.
Sensitivity Classifier
Automatically detects and classifies sensitive data — credentials, PII, API keys, and secrets.
Fragmentation Detector
Detects memory fragmentation attacks that try to spread malicious payloads across entries.
Credential Leak Detection
25+ patterns across 11 providers plus entropy analysis — catches API keys, tokens, and secrets before they reach memory.
Sub-Agent Activity Dashboard
Complete visibility into what your AI sub-agents are doing — in real time
Agent Registry
Track every sub-agent: identity, trust level, and permission scope
Trust Timeline
Visualise trust score changes over time with full history
Flagged Operations
Instantly spot suspicious memory operations and credential access
Real-Time Alerts
Live alert feed for security events across your entire agent fleet
Iron Dome
Agent-aware security that controls what your AI can do, not just what it remembers
Injection Scanning
Deep pattern matching for prompt injection, jailbreaks, and social engineering in any text
Action Gates
Control which actions agents can take — send emails, delete files, call APIs — with approval rules
Security Profiles
Pre-built profiles — school, enterprise, personal, paranoid — or customise your own
Full Audit Trail
Every scan, block, and quarantine is logged with forensic detail for compliance
Works With Any Memory System
Use ShieldCortex as a standalone memory system or as a security layer for any memory backend — OpenClaw, LangChain, custom systems, or anything else
OpenClaw Hook
Native integration with OpenClaw — opt-in auto-memory with smart deduplication and novelty filtering
Universal Memory Bridge
Guard any memory backend with the full defence pipeline — every read and write scanned automatically
Python SDK
Full Python support via pip — same scanning, same protection, for Python-based agents and pipelines
Hierarchical Trust Security
Running multiple agents? ShieldCortex prevents rogue sub-agents from accessing sensitive data with automatic trust decay.
Credential Isolation
RESTRICTED memories blocked below trust 0.7 — sub-agents can't access your API keys or secrets
Trust Decay
Each sub-agent level reduces trust: user (0.9) → task (0.63) → subtask (0.44)
Depth Circuit Breaker
Agents beyond depth 5 get trust = 0. Automatic protection against runaway chains
Auto-Quarantine
Low-trust agents write to quarantine for human review before storage
Trust ≥0.7 → Read all, write direct | Trust 0.5–0.7 → Limited read, quarantine writes | Trust <0.5 → Own data only
Simple Pricing
Complete memory system + security layer. Free forever for open source.
Free
- ✓ Persistent SQLite storage
- ✓ Semantic search
- ✓ Project scoping
- ✓ Auto-decay & importance
- ✓ MCP native support
- ✓ Memory Firewall
- ✓ Audit Logger
- ✓ Trust Scorer
- ✓ Multi-Agent Defence
Pro
- ✓ Memory consolidation
- ✓ Contradiction detection
- ✓ Knowledge graphs
- ✓ Iron Dome behavioural protection
- ✓ Credential leak detection
- ✓ Dashboard & Alerts
- ✓ Priority email support
Team
- ✓ Sub-Agent Activity Dashboard
- ✓ Universal Memory Bridge
- ✓ Custom Iron Dome patterns
- ✓ RBAC & team invites
- ✓ Priority support
Enterprise
- ✓ Self-hosted deployment
- ✓ SIEM integration
- ✓ Compliance exports
- ✓ Custom defence rules
- ✓ Dedicated support + SLA
All plans include: local defence pipeline · npm updates · macOS, Linux & Windows · MIT licence
Iron Dome
ShieldCortex protects what your agent remembers. Iron Dome protects what it does. Six layers of behavioural security — instruction gateway, injection scanning, action gating, PII protection, kill switch, and full audit trail.
Works With Your Stack
Drop-in security for the tools you already use
Secure Your AI Agent Memory
Open source, MIT licensed, free forever. Get started in under a minute.