Lyrie 3.1.0: Open-Source Autonomous Pentesting Agent Adds Memory Encryption and Agent Trust Protocol

OTT Cybersecurity has released version 3.1.0 of Lyrie, an open-source autonomous penetration testing agent that compresses weeks of manual security assessment into a single command-line workflow. The tool, available for free on GitHub, now includes XChaCha20-Poly1305 memory encryption for protecting sensitive threat data in transit, seven new proof-of-concept generators, and three new deep scanners targeting Rust analysis, taint engine processing, and AI-driven code review.

Lyrie splits into two installable packages: lyrie-omega, a Python CLI that handles scanning, pentesting, and red-teaming, and @lyrie/atp, a TypeScript and Node.js SDK implementing the Agent Trust Protocol (ATP). The core pentest workflow, triggered by the command `lyrie hack`, runs a seven-phase pipeline: reconnaissance, fingerprinting, scanning, exploitation, proof-of-concept generation, and report output. The tool targets live URLs and local source trees and outputs findings in SARIF format for direct integration with GitHub Code Scanning.

The new proof-of-concept generators cover prompt injection, authentication bypass, CSRF, open redirect, race conditions, secret exposure, and cross-site execution. The AI red-teaming module supports five attack strategies against LLM endpoints, including gradient-based suffix attacks that require H200 GPU infrastructure. The repository now ships 25 tested commands spanning core security operations, binary analysis, governance, and self-improvement workflows.

The Agent Trust Protocol addresses a critical gap in how autonomous AI agents authenticate themselves and communicate scope to the systems they interact with. Enterprises deploying agents that send email, execute code, or authorize transactions have had no standard mechanism for verifying agent identity or checking whether an agent's instructions have been tampered with. ATP uses Ed25519 signatures and supports delegation chains, revocation lists, and multisig configurations. A verifying system can confirm in real time who the agent is, what it is authorized to do, and whether its authority has been revoked. The specification carries 143 passing tests and is slated for submission to the Internet Engineering Task Force.

Lyrie's release comes amid a broader industry shift toward autonomous security tools. The rise of AI-driven vulnerability discovery and automated penetration testing is fundamentally altering the threat landscape, rendering traditional security-through-obscurity strategies obsolete. However, the same technology that enables defenders to find bugs faster also arms attackers with more sophisticated tools, creating a 'perfect storm' for security teams.

Lyrie is available for free on GitHub, and OTT Cybersecurity has published the entire codebase to encourage community contributions and transparency. The tool's modular architecture and cryptographic identity layer position it as a potential standard for autonomous security agent operations, though its effectiveness will depend on widespread adoption and continued development.