Open-source sandboxed agent harness for teams. Giving every employee a secured personal agent.

Playwright for coding agents. Benchmark Claude Code, Codex, Gemini, and OpenCode on your own tasks - and test that your skills, MCP servers, and CLIs work when an agent uses them. Sandboxed YAML suites, activation checks, A/B experiments, CI gates.

Containment for AI agents - user isolation, sandboxed execution, network controls, backup/rollback. TLA+ verified.

Claude Code, Codex, and multi-agent configuration system: principles, hooks, skills, and workflow patterns for AI-assisted development

Comprehensive sets of standards and practices designed to elevate the capabilities of AI coding agents.

A ruleset that turns AI coding assistants (e.g. Claude Code) into disciplined Linux, FreeBSD & macOS sysadmins. Manages servers via SSH and localhost with safety guardrails, checklists, and team support. Named after the Heinzelmännchen — helpful kobolds who do the housework while you sleep.

Safe local execution layer for AI agent tools. Build, validate, and publish MCP tools with a no-pass-no-run workflow — cross-platform desktop app powered by Spring AI.

Local-first runtime for project-scoped AI coding-agent sessions, with durable state, authority boundaries, and multi-harness interoperability.

Vigilante is a sandbox-first orchestration layer for coding agents. It isolates every task in a git worktree, enforces strict credential scoping, and gives you full audit logs — so your agents can't burn down production.

MCP Server Security Standard (MSSS): an open, testable security control standard for certifying MCP servers, with levels, evidence requirements, and reporting schemas.