At Databricks, we use and construct brokers extensively, from coding with them at scale to transport agent merchandise like Genie. However though the capabilities of brokers have gotten significantly better, working with them feels clunky. As customers, we frequently have 4-5 brokers open directly (coding brokers, Gemini search, and so forth) and spend our time copy-pasting textual content between them and Docs, Slack, and different collaboration instruments. And as agent builders, we’re on a treadmill to enhance our brokers by combining the most recent harnesses, SDKs and fashions. The issue is that LLM capabilities are wrapped into an agent harness, and these harnesses have totally different interfaces that make combining them or swapping them tough.
So we constructed Omnigent: a meta-harness that sits above the brokers you already use (Claude Code, Codex, Pi, or customized brokers) and makes them interoperable elements of a richer system. Omnigent targets the issues the place a single harness stops: it provides straightforward methods to compose a number of brokers, management them with superior insurance policies, and collaborate dwell with teammates.
We consider individuals will quickly work with brokers via this new layer, the meta-harness. That’s why at present we’re open sourcing Omnigent below Apache 2.0.
Why construct a meta-harness?
At Databricks, we adopted coding brokers early throughout our 5000+ member engineering workforce and constructed hundreds of brokers for patrons. That have satisfied us that the frontier of agent engineering is transferring up a stage. The perfect outcomes now not come from a single mannequin in a single harness: Harvey beat a frontier mannequin on high quality and value by giving an open-source employee mannequin a frontier advisor it may possibly name, Anthropic constructed its analysis product as a lead agent orchestrating parallel subagents, and our personal Genie makes use of totally different LLMs for planning, search, and code era. Engineers are altering how they work, too: as a substitute of prompting one agent at a time, they design loops that drive entire groups of brokers.
These patterns span a number of harnesses, fashions, and folks, however every harness solely understands its personal classes. To mix brokers, govern them, and work on them with different individuals, you want a layer above the harness. Omnigent is that layer, and it offers:
- Composition. Mix a number of fashions, harnesses, and strategies with out rewriting code, and swap between Claude Code, Codex, Pi, and your individual brokers with one-line adjustments.
- Management. Stateful, contextual insurance policies that monitor agent actions and implement guardrails like value budgets and permissions on the meta-harness layer, not by way of prompts.
- Collaboration. Share dwell agent classes by way of URL and evaluate recordsdata in them collectively, so teammates can evaluate, remark, and steer brokers collectively in actual time.
How Omnigent works
Omnigent introduces a standard interface above command-line brokers and agent SDKs to allow you to simply mix and interchange them, after which focuses on the shared issues the place a harness stops. The important thing perception is that nonetheless every agent harness calls into its LLM internally, the interface to customers is identical: messages and recordsdata in, textual content streams and power calls out. Thus we constructed a standard API that wraps each terminal-based coding brokers (Claude Code, Codex, Pi, and so forth) and SDKs (OpenAI Brokers, Claude Brokers SDK, and so forth).
On prime of this interface, the present model of Omnigent provides the next key options:
- Actual-time collaboration: you possibly can invite different individuals to view your agent session, touch upon recordsdata in its workspace, and even ship instructions, so your classes and dealing directories grow to be the principle place you collaborate.
- A number of interfaces to the identical agent: when you join an agent akin to Claude Code to the Omnigent server, you possibly can entry it on the net, cellular, Mac OS native app, or APIs.
- Cloud execution: launch any agent by yourself machine or on hosted sandbox suppliers like Modal and Daytona, for secure collaboration in a airtight surroundings.
- Contextual safety insurance policies: Omingent’s safety insurance policies transcend the easy “enable X / deny Y” of coding brokers, to trace dynamic state about every session and make smarter selections. For instance, you possibly can say that after an agent downloads a brand new bundle from npm, it ought to require human approval to git push, or that it ought to solely be capable of write to docs it created, not any doc.
- Price insurance policies: One of many issues we monitor dynamically is every session’s LLM value. For instance, you possibly can ask Omnigent to pause an agent and ask to proceed after each $100 it spends.
- Sturdy OS sandbox: In Omnigent, we embrace a versatile OS sandbox from our safety workforce with the flexibility to flexibility lock down OS entry and intercept and remodel community requests (e.g., don’t let an agent ever see your GitHub safety token, however as a substitute, inject it solely within the egress proxy on accredited requests).
- Multi-harness authoring: Specify a customized agent as a YAML and port it throughout harnesses with a one-line change, or mix subagents utilizing totally different harnesses in the identical agent.
These options are simply scratching the floor of what may be completed on the meta-harness layer, nonetheless, and we anticipate to see much more concepts quickly from our workforce and the open supply group. Some objects on our roadmap embrace automated optimization on the meta-harness stage with GEPA, code-based introspection inside brokers just like MemEx and RLM, an Omnigent Server MCP so brokers can work throughout your classes, and extra harnesses. We’ve additionally made Omnigent straightforward to deploy on a variety of infrastructure, together with Fly.io, Railway, Modal and Daytona sandboxes, and lots of LLM suppliers, and we welcome patches for extra integrations.
A brand new layer for working with brokers
Lots of the largest shifts in our trade got here from transferring to a brand new layer of abstraction: for instance, whereas engineers used to handle particular person processes and servers, they’ll now handle a complete fleet by way of cloud programs like Kubernetes and Terraform.
We predict brokers are on the similar level at present. Every harness is its personal silo, with its personal context, its personal controls, and its personal means of operating, and none of it carries over while you swap instruments. Furthermore, many issues intrinsically span harnesses, together with composition, safety and collaboration. A meta-harness lifts your work above any single harness, so your classes, insurance policies, and expertise stick with you irrespective of which agent or mannequin is operating. The fashions and harnesses will preserve altering as the sphere evolves; the layer you’re employed at should not should.
We’re constructing that layer within the open, and we would love so that you can construct it with us.
Attempt it out
Omnigent is open supply in alpha at present.