rossoctl/rossoctl

Main rossoctl repo - installer, UI and docs

Rossoctl – Platform primitives for trustworthy AI agents

What it is – Rossoctl is an open‑source, framework‑neutral platform that sits between an AI agent and the external resources it uses (models, tools, users, other agents). It provides a data‑plane called RossoCortex that intercepts all agent interactions and enforces security, identity, authorization, resilience and observability guarantees. On top of the data‑plane Rossoctl ships a set of reusable services (skills, tools, memory, knowledge base, sandboxes) and tooling for administration, monitoring and governance.

Why it matters – Modern LLM‑driven agents are dynamic: they pick tools at runtime, modify their own prompts, and can drift from their original goals. Existing cloud runtimes (e.g., Kubernetes) give such guarantees to traditional micro‑services, but agents lack a comparable safety net. Rossoctl fills that gap by decoupling agent logic from operational guarantees so that any agent—regardless of the underlying framework—gets the same admission control, isolation, failure‑recovery and auditability.

Core components

Component Role Maturity (as of v0.7.0)
RossoCortex (data plane) Transparent intercept that validates identity, authorizes actions, can compact context, analyse data‑flow, and recover from failures. Identity & Authorization – Ready; Intent‑based access, tool validation, context compaction, data‑flow analysis, failure recovery, user interaction – beta/alpha
Rossoctl services Building‑block APIs that agents call:
  • Skills – versioned, governed capabilities
  • Tools – external utilities
  • Memory – persistent state
  • Knowledge base – searchable facts
  • Sandboxes – isolated execution environments | Generally available; specific features evolve with the platform | | Tooling | CLI, UI, CRDs, observability dashboards, security & governance helpers. | GA (UI, CLI) |

How you interact

  • SDK – language bindings for agents to talk to RossoCortex.
  • Hooks – side‑car containers or init scripts that inject the intercept.
  • Gateway – a network entry point that proxies agent traffic.
  • Orchestration layer – Kubernetes operators/CRDs that manage the whole stack.

Typical workflow (quick‑start)

  1. Prerequisites – Python ≥ 3.9 (with uv), Docker/Podman, a local K8s cluster (kind/Rancher Desktop), and Ollama for local LLM inference.
  2. Deploy – Clone the repo, check out the latest release tag (e.g., v0.7.0), configure secrets, then run the provided setup-rossoctl.sh script which installs the Helm chart, optional UI, SPIRE identity provider, sandbox runtime, and build pipelines.
  3. Access UI – The script prints service URLs and credentials; open http://rossoctl-ui.localtest.me:8080 to log in.
  4. Use the UI – Import an A2A‑compatible agent, deploy MCP tools, run interactive chats, and view trace/network visualisations.
  5. Extend – Bring your own agent (any framework) via the Bring your own agent guide, or develop custom skills/tools that are governed by Rossoctl.

Who should consider Rossoctl?

  • Enterprises building production‑grade LLM agents that need audit trails, role‑based access, and guaranteed recovery.
  • ML Ops teams that want a Kubernetes‑native way to run agents alongside traditional workloads.
  • Security‑focused developers looking for verifiable agent identities and policy‑driven tool usage.
  • Researchers experimenting with new agent‑to‑agent (A2A) protocols or model‑context (MCP) integrations while still having a sandboxed testbed.

Documentation & community

  • Full docs – https://rossoctl.dev/docs/ (quick‑starts, concepts, architecture, security, reference).
  • Slack community – join via the badge in the README.
  • OpenSSF Scorecard – shows the project’s supply‑chain health.
  • Contribution guide – CONTRIBUTING.md outlines how to submit PRs, run CI, and add demos.

License

Apache 2.0 – permissive open‑source license allowing commercial use and modification.

Bottom line – Rossoctl provides the missing “operating system” layer for AI agents, giving them the same production guarantees that Kubernetes gives to containerized services, while remaining framework‑agnostic and built on open standards (A2A, MCP).

Related

  • Project
    Xueyang-Song/paper-pilot

    Paper Pilot is a desktop (Electron + React) research assistant that crawls 8+ scholarly databases, downloads PDFs, indexes them with SQLite + FTS5 + vector search, and lets you ask a local or hosted LLM for evidence‑grounded answers. All data stays on your machine; the app provides audit‑ready citation traces and a reproducible workflow for literature review.

  • Project
    usnistgov/dioptra

    Dioptra is NIST’s open‑source Docker‑based platform for testing AI models against trustworthiness criteria (fairness, safety, explainability, etc.). It offers a REST API, web UI, and Python client, lets you define experiments in YAML, supports plugins for PyTorch/TensorFlow, and tracks reproducible, traceable runs for developers, auditors, researchers, and red‑teamers.

  • Project
    Prorise-cool/Claude-Code-Multi-Agent

    Claude Code Multi‑Agent is an open‑source framework that gives Anthropic’s Claude Code project awareness and expert‑agent capabilities. It uses a Python Hook system and a locally‑run Ollama model to automatically detect the codebase type, analyse user intent, load 300+ domain‑specific “Skills” (backend, frontend, testing, security, etc.), and enforce automatic documentation (DEVELOPMENT.md, KNOWLEDGE.md, CHANGELOG.md). After cloning, installing Ollama, a model, and uv, you drop your project into the repo and Claude Code instantly knows its context, suggests appropriate tools, and can be driven with slash commands like `/backend‑specialist` or workflow commands such as `/kiro/spec`. Key points: - Zero‑config start; all configuration lives in `.claude/settings.json` and `prompts.json`. - Hooks run on session start, user prompt, tool use, etc., handling project detection, intent analysis, and doc updates. - Skills are defined by `SKILL.md` files and are auto‑discovered. - Works entirely locally (no external API), integrates with Git, and is MIT‑licensed. Ideal for developers or small teams who want a reproducible, on‑premise AI coding assistant.

  • Project
    Stability-AI/stable-audio-metrics

    A collection of evaluation metrics for music and audio generative models, providing tools to calculate Fréchet Distance, KL divergence, and CLAP scores for long-form stereo audio.