PIC Standard is an open-source, local-first check that runs right before an AI agent uses a tool. Before a payment, data export, deletion or command goes through, the agent must show what it wants to do, where its information came from, and the evidence behind it, such as a...
Pipelock is an open-source agent firewall written in Go. It runs as a sidecar or local service between an agent and the network, scanning HTTP, WebSocket, and Model Context Protocol traffic through an 11-layer pipeline.
United States
Self-hosted open-source bastion for privileged SSH and RDP access from the browser, with session recording and access control.
Warpgate Bastion
HOL Guard is an open-source, local-first runtime security layer for AI agents and automation. It sits between any AI harness and the tools it wants to run, pausing risky package installs, secret reads, shell commands, and MCP changes for review before execution.
Marshall Islands
Security API for AI agents to detect prompt injection, jailbreak attempts, and unsafe operational intent before execution.
Cloudflare
Tragentics is the AI agent security platform that authenticates every agent, injects keys from an encrypted Credential Vault so agents never hold them, routes every call through a content-blind relay, and records a metadata-only audit trail — across platforms and protocols.
United States
AI Agent Security Testing — 112 attacks across 14 categories. Prompt injection, jailbreaks, MCP poisoning, agency hijacking & more. Test any AI agent in 5 minutes.
