OneCLI is an open-source credential gateway that stores API keys in a secure vault and injects them into AI agents, allowing agents to make HTTP calls without exposing keys. It uses a Rust-based gateway, a Next.js web dashboard, and AES-256-GCM encryption to manage credentials and permissions. OneCLI provides a quick start guide for local installation and deployment. AI summary
Firehose
Filtered to Hacker News, tagged “security” · clear filters
Browse: People · Companies · Papers · Podcasts · Hacker News
Browse by tag
A poorly designed security harness can execute scripts, misleadingly labeling it as a "rogue AI" scenario, rather than a flaw in the security implementation. This highlights a critical issue in AI development where security is often overlooked. The term "rogue AI" is often misused to describe legitimate security vulnerabilities. AI summary
Detecting ANSI Escape Sequence Injection (AESI) in MCP servers is crucial, as attackers can hide instructions from humans while leaving them legible to AI models. DAST (Dynamic Application Security Testing) is the natural approach to catching this vulnerability, as it exercises a running target from the outside and inspects real responses for evidence of a vulnerability. Two variants of the attack, direct-fetch and stored AESI, put agent actions, human-in-the-loop bypass, and log manipulation at risk. AI summary