https://cybertoolchain.io/tools/object-oriented-agents New releases and features in Object-Oriented Agents, tracked by The Cyber Toolchain. en Thu, 30 Jul 2026 06:18:02 GMT https://github.com/NVIDIA-NeMo/labs-OO-Agents/releases/tag/v0.0.7 https://github.com/NVIDIA-NeMo/labs-OO-Agents/releases/tag/v0.0.7 Thu, 30 Jul 2026 06:18:02 GMT OO Agents v0.0.7 adds a NOOA cybergym agent example, viewer ingest resource limits, and secret redaction from telemetry. • Constrains playground custom-model endpoint and `api_key_env` to server-declared pairs, preventing clients from supplying arbitrary LLM endpoints. • Adds resource limits on the viewer ingest path to cap memory and request size for the trace-viewer API. • Requires an auth token and restricts CORS on the viewer API, narrowing the attack surface of the local trace viewer. • Confines `ShellTools` file operations within the current working directory, limiting the filesystem surface available to agent-executed shell commands. • Adds a NOOA cybergym agent example (`examples/`) demonstrating a cybersecurity-oriented agent built on the OO Agents framework. • Replaces eval() in the `CodeAct` constructor-string coercion path with an AST decoder, removing arbitrary code execution from that code path. • Publishes the package to PyPI via Trusted Publishing on GitHub Releases, enabling `pip install nooa` / `uv add nooa` installs from official releases. Full history on Cyber Toolchain: https://cybertoolchain.io/tools/object-oriented-agents