Modulo With Your Existing Agent Runtimes
Plenty of teams already run agents. The agents live in a service in AWS, or they run as a job in CircleCI, or they are scripts on a build host. The orchestration and the agent are not the missing piece. The governance and the record are.
Modulo plugs into the runtime that exists.
Typed contracts make the runtime swappable
A Modulo agent can declare an input schema and an output schema. Where a schema is declared, the input is validated before dispatch and the output is validated before any gate or reviewer sees the result. Any runtime that respects the contract can fill a node, which is what makes the orchestrator agent-agnostic: the pipeline does not know or care whether the work happens in a Modulo sandbox or a service in AWS. See How input/output schema works with external agents.
Connect to the CI that runs the agents
Modulo ships CI-runner connectors for CircleCI, GitHub Actions, GitLab CI, Buildkite, Jenkins, TeamCity, and Azure Pipelines. Each one starts a run, reads its status, reads its logs, and lists runs. Wiring a pipeline node to dispatch to these runners is in development. See Connectors.
An agent runtime that exposes an API and is not one of the above is reachable through the generic HTTP adapter or a filesystem or shell connector bound to a restricted base path. An agent service running in AWS is wired in through the generic HTTP connector, or by calling back into Modulo through a webhook trigger.
Bring work back in
Two triggers close the loop. A webhook trigger accepts an inbound POST with HMAC validation and maps the payload onto the entry agent's input schema, so your runtime can hand work back to a governed pipeline. An agent-signal trigger fires a pipeline from another pipeline's output. See Triggers for both, plus cron and polling.
The record follows the work
Whichever path the agent takes, the run writes the same audit events and the same versioned snapshot: the trigger, the dispatch, the validated output, the gate decision, and the final artifact. The runtime you already operate gains a governance layer and a queryable record without being rebuilt.
Where to go next
- How input/output schema works with external agents: the contract in detail
- Connectors: CI runners, HTTP, filesystem, and shell
- Pipelines: how external steps compose with gates and evals
Related use cases
AI Agent Cost Visibility
Per-run and per-team token cost, spend limits, anomaly alerts, and hard ceilings for AI agents, from the same place the run is governed.
Audit Trail for AI-Generated Code
A hash-chained, append-only audit trail plus a versioned snapshot for every run, so you can prove who approved what an AI agent changed.
What Modulo Does With GitHub
The GitHub connector's read and write operations, GitHub Actions triggering, and how Modulo gates agent pull requests without replacing branch protection or CODEOWNERS.