A system-level defense framework for AI agents that intercepts and blocks indirect prompt injection attacks using strict instruction provenance and dynamic policy enforcement.
Inspired from https://github.com/anthropic-experimental/sandbox-runtime. A python sandboxing tool for enforcing filesystem and network restrictions on arbitrary processes at the OS level, without requiring a container.
Secure operating environment for bounded agent execution with signed plans, SATL envelopes, ToolGate enforcement, and audit-ready controls.
To enable teachers and students to use agentic AI for free. A lightweight, Docker-free environment for running Hermes Agent inside GitHub Codespaces, with remote access from the Hermes Desktop App. WARNINGS: You will need to anonymise student names before uploading any data. You will need to be tech-savvy to use effectively.
To associate your repository with the secure-agents topic, visit your repo's landing page and select "manage topics."