AgentIndex icon
AgentIndex
norviq logo
norviq
★ 26
vs
www-project-agent-memory-guard logo
www-project-agent-memory-guard
★ 188

norviq vs www-project-agent-memory-guard

norviq: Norviq is a policy enforcement point that intercepts and evaluates every LLM agent tool call against OPA/Rego policies on Kubernetes. It prevents potential risks like prompt injection or reasoning errors by ensuring tools only run if explicitly allowed by policy, enhancing AI safety and auditability.; www-project-agent-memory-guard: OWASP Agent Memory Guard is a runtime defense solution designed to prevent AI agents from being weaponized through their own memory by detecting and blocking memory poisoning attacks. It provides a robust security layer that intercepts malicious writes to agent memory, ensuring the integrity and safety of AI operations even after context resets.

01

TL;DR

norviq logo
norviqSecurity & Safety · ★ 26

Preventing LLM agents from executing unauthorized or dangerous tool calls, such as `execute_sql` or `send_email` without explicit policy approval.

Read norviq full review →
www-project-agent-memory-guard logo
www-project-agent-memory-guardSecurity & Safety · ★ 188

Reference implementation for OWASP ASI06: Memory Poisoning.

Read www-project-agent-memory-guard full review →
02

Side-by-Side Comparison

Field
norviq logonorviq
www-project-agent-memory-guard logowww-project-agent-memory-guard
Category
Security & Safety
Security & Safety
Stars
★ 26
★ 188
License
Apache-2.0
Apache-2.0
Install Command
kubectl create namespace norviq && kubectl create namespace default && helm install norviq oci://ghcr.io/norviq-dev/charts/norviq --version 0.2.5 -n norviq --set 'policyQuotaNamespaces={default}' --set config.dbSslMode=disable
pip install agent-memory-guard
Updated
3w ago
1d ago
Open Source
Yes
Yes
Website
↗ Visit
↗ Visit
GitHub
↗ GitHub
↗ GitHub
Tags
LLM Agents, Policy Enforcement, Kubernetes
AI Security, Agent Defense, Memory Protection
03

Features

norviq logonorviq
01Policy enforcement using OPA/Rego for sub-second, fail-closed evaluation of agent tool calls.
02Kubernetes-native deployment with custom resource definitions (CRDs), a mutating webhook, and Helm chart support.
03Supports workload identity using SPIFFE/SPIRE SVIDs for granular, per-agent-class and per-namespace policy scoping.
04Provides a Console UI for policy catalog/editor, attack/asset graphs, agent trust management, and audit stream.
05Includes a red-team suite with built-in adversarial tests to validate policy effectiveness against attacks like prompt injection or SQLi.
www-project-agent-memory-guard logowww-project-agent-memory-guard
01Integrity: SHA-256 baselines flag out-of-band tampering with immutable keys.
02Threat detection: Built-in detectors for prompt injection, sensitive data leakage, protected-key modifications, size anomalies, and self-reinforcement loops.
03Policy enforcement: YAML-defined rules map findings to actions like `allow`, `redact`, `quarantine`, or `block`.
04Forensics: Every decision emits a structured `SecurityEvent`, with point-in-time snapshots for rollback to a known-good state.
05Drop-in middleware: Ships with `GuardedChatMessageHistory` for LangChain and general framework-agnostic `MemoryStore` protocol.
04

Use Cases

norviq logonorviq
↳Preventing LLM agents from executing unauthorized or dangerous tool calls, such as `execute_sql` or `send_email` without explicit policy approval.
↳Ensuring compliance by enforcing policy-driven gates on agent tool interactions and providing comprehensive audit logs for all decisions.
↳Implementing fine-grained security policies for different agent classes and namespaces within a Kubernetes environment.
↳Detecting and blocking prompt injection or reasoning errors that could lead to malicious tool execution or data exfiltration.
www-project-agent-memory-guard logowww-project-agent-memory-guard
↳Reference implementation for OWASP ASI06: Memory Poisoning.
↳AI security research, such as at Microsoft.
↳Securing multi-tenant agent deployments with compliance requirements in enterprise teams.
↳Protecting LangChain, OpenAI Agents SDK, AutoGen, mem0, and CrewAI based agents from memory-based attacks.
FAQ

FAQ

What is the difference between norviq and www-project-agent-memory-guard?
Both norviq and www-project-agent-memory-guard are in the Security & Safety category. norviq has 26 stars, while www-project-agent-memory-guard has 188 stars.
Which is better, norviq or www-project-agent-memory-guard?
The best choice depends on your use case. Choose norviq if Preventing LLM agents from executing unauthorized or dangerous tool calls, such as `execute_sql` or `send_email` without explicit policy approval., and www-project-agent-memory-guard if Reference implementation for OWASP ASI06: Memory Poisoning..
Is norviq free or open source?
Yes, norviq is open source on GitHub (Apache-2.0).
Is www-project-agent-memory-guard free or open source?
Yes, www-project-agent-memory-guard is open source on GitHub (Apache-2.0).
Cloud Infrastructure
Ad

Deploy on DigitalOcean

Host your MCP servers, Docker containers, and AI agents with $200 free credit over 60 days.

Claim $200 Free Credit
→

Related

Alternatives to norviq →Alternatives to www-project-agent-memory-guard →norviq details →www-project-agent-memory-guard details →