AgentScout / By role / Security engineers

AI agents for security engineers (2026)

Pentest agents, guardrails, identity and sandboxing tools for securing apps and the agents that touch them. We picked the top tools for 5 jobs security engineers do every day: 24 tools in all, 20 of them free or with a free tier.

Jump to: Security testing · Agent security and guardrails · Smart contract security · Agent identity and logins · Sandboxes

Free tools from AgentScout for security engineers: Agent setup generator (guardrail rules in CLAUDE.md, AGENTS.md)

How we picked: each section takes the tools matched to that task in our directory, putting tools whose category and stated use cases fit the task first, then ordering by tracked GitHub stars. No paid placements.

Security testing

Agents that probe your app, code and smart contracts the way an attacker would: running penetration tests, red-teaming AI systems and flagging vulnerabilities with suggested fixes. All 8 tools →

Strix Open source + cloud
Open-source AI agents that run penetration tests on your app to find and help fix security vulnerabilities. · ★ 69,022
PentAGI Open source
Autonomous multi-agent system for running penetration testing tasks. · ★ 30,728
AI-Infra-Guard Open source
Tencent's AI red-teaming platform that scans agents, skills and MCP servers for risks. · ★ 6,887
PentestAgent Open source
Agent framework for black-box security testing and bug bounty work. · ★ 3,188
SonarQube MCP Server Open source
SonarQube MCP server for code quality and security analysis from agents. · ★ 665
Almanax See site
AI security engineer that finds vulnerabilities in applications and smart contracts and reviews pull requests; acquired by depthfirst (June 2026).

Agent security and guardrails

Guardrails, policies and monitoring that keep agents from doing what they shouldn't. All 31 tools →

Strix Open source + cloud
Open-source AI agents that run penetration tests on your app to find and help fix security vulnerabilities. · ★ 69,022
OpenSandbox Open source
Secure, fast sandbox runtime for AI agents to run code. · ★ 15,907
CubeSandbox Open source
Lightweight, concurrent sandbox for running AI agent code securely. · ★ 12,960
Guardrails AI Open source + cloud
Open-source framework for validating LLM inputs and outputs with configurable guardrails. · ★ 7,529
Agent Governance Toolkit Open source
Microsoft toolkit for policy enforcement, identity and sandboxing for AI agents. · ★ 6,435
Failproof AI Open source
Observability and policy enforcement for AI agent harnesses. · ★ 5,560

Smart contract security

AI auditors and agents that review smart contracts for exploitable bugs before and after deployment, plus tools that help build and check Web3 code. All 5 tools →

Almanax See site
AI security engineer that finds vulnerabilities in applications and smart contracts and reviews pull requests; acquired by depthfirst (June 2026).
ChainGPT See site
Suite of crypto AI tools: chatbot, smart-contract generator and auditor.
Nora See site
Deep-reasoning AI coding agent for building and securing Web3 apps and smart contracts.
Octane See site
Agentic security analysis that finds exploitable vulnerabilities in mission-critical code, including DeFi smart contracts.
Sherlock AI See site
Sherlock's AI auditor for smart-contract vulnerabilities.

Agent identity and logins

Identity, auth and credential tools that let agents sign in to apps, prove who they act for and use secrets without ever seeing the raw passwords or keys. All 10 tools →

Casdoor Open source
Agent-first identity and access management with an MCP and agent gateway. · ★ 14,592
Bindu Open source
Identity, communication and payments layer that lets AI agents talk to and pay each other. · ★ 10,056
Agent Governance Toolkit Open source
Microsoft toolkit for policy enforcement, identity and sandboxing for AI agents. · ★ 6,435
Open Connector Open source
Open-source auth gateway connecting SaaS apps to AI agents via SDK, CLI and MCP. · ★ 6,031
OneCLI Open source
Sandboxed agent harness for teams with secure credential handling. · ★ 3,576
Composio Freemium
Tool integrations and auth for agents across hundreds of apps.

Sandboxes

Sandboxes that let agents run code and tools in isolation, away from your machine and data. All 22 tools →

Context Mode Open source
Context window optimizer for coding agents that sandboxes tool output and persists session memory. · ★ 30,192
Cua Open source + cloud
Open-source infrastructure for computer-use agents: sandboxes, drivers across operating systems and benchmarks. · ★ 29,966
NVIDIA OpenShell Open source
Sandboxed runtime for autonomous agents with declarative YAML policies on file, network and credential access. · ★ 16,156
OpenSandbox Open source
Secure, fast sandbox runtime for AI agents to run code. · ★ 15,907
CubeSandbox Open source
Lightweight, concurrent sandbox for running AI agent code securely. · ★ 12,960
Open SWE Open source
LangChain's asynchronous coding agent that takes work from GitHub issues, Slack or Linear, implements it in an isolated sandbox and opens a pull request. · ★ 11,034