Posted Oct 30, each runtime session gets an authenticated URL that forwards HTTP, SSE and WebSocket traffic to 1–5 allowlisted ports (8000–12000) and stops working when the session ends. It's microVM runtimes only, with public HTTPS and no PrivateLink for inbound connections.
Shared on HN Oct 31, the zero dependency agent reads and edits files, runs shell and git with checkpoints and undo, speaks MCP, loads SKILL.md skills and works with any model in the models.dev catalog, using about 1.7 MB of memory on Windows. Only Windows is fully tested (Linux builds and runs, macOS is source only), and its source available license bans commercial use and public forks…
Released Nov 1 under Nace.AI Open RAIL M, Drex Max returns a probability for every option in one forward pass instead of generating text, so routing, triage and guardrail checks can run locally: a 4 bit build fits 24 GB Apple Silicon Macs, with a Q8 0 GGUF and llama.cpp and Ollama forks for 48 GB machines or CPU. Its 67.52 on Decision Index 0.4 is self reported, and local Mac and CPU paths are text only for now…
Launched Oct 30, claude mcp add transport http scope user deployhq https://mcp.deployhq.com/mcp (or codex mcp add deployhq url ... ) signs you in through the browser, with no package or API key, and six tools answer questions about your projects, servers and recent deployments. It can't deploy, roll back, read logs or see secrets yet, needs a plan with API access, and admins can switch it off account wide…
Released Nov 1 with no listed breaking changes, /autofix pr now hands a check that's still failing after three fix pushes to a human and comments on the PR when it's blocked, while the PR watch skill gains in place to work in your own checkout. Generated UI cards are on by default, and an experimental Helm chart deploys the Managed Agent in one command…
Add it with claude mcp add transport http mitigant https://mcp.mitigant.io/mcp plus an API key, and your agent can pick from 111 MITRE ATT&CK mapped attacks on AWS, Azure and GCP, run them inside the access you grant, collect evidence and reverse the changes afterwards. These are real actions on live infrastructure, so it's meant for red teaming and detection checks on accounts you own (MCP page)…
Spotted something we missed? Start a thread.