Lorikeet Security: The AI pentester whose findings are countersigned by humans

Lory is an autonomous AI pentester. Give her a signed scope and she runs recon, chains exploits, and drafts findings with evidence. A human pentester countersigns every finding, and each run shows what she tested and what she didn't reach. topics: Security, Artificial Intelligence, Developer Tools, SaaS pricing: Paid (pay-as-you-go credits, no subscription) first comment (maker comment): Hey Product Hunt! I'm [name] from Lorikeet Security. We built Lory because AI security tools have a noise problem. Most of them hand you a pile of "possible" vulnerabilities and leave you to figure out which ones are real. Lory works differently. You give her a signed scope, and she runs recon, picks her own attack vectors, validates exploits, chains them together, and drafts the finding with evidence. Then it goes into a review queue that even you can't see. Nothing reaches your backlog, your auditors, or your devs until one of our human pentesters has read the evidence and countersigned it. That gate isn't a setting, it's built into how the product works. The other thing we're proud of: every engagement publishes the attack vectors Lory planned, the ones she ran, and the ones she never reached. Most pentest reports quietly imply they covered everything. Ours tell you exactly where to go deeper next time, and you can hand that coverage record straight to an auditor. A few other things: She runs 57 attack playbooks across web, API, cloud, and AI systems Findings stream into your portal in real time and can route to Slack, Jira, and your repos There's an open-source in-network agent, a CLI, and a VS Code extension on our GitHub No subscription, just load credits and run an engagement Would love feedback, especially from anyone who's been burned by noisy scanners or AI tools before. What would it take for you to trust an AI agent inside your network?