A standalone runtime for portable agents

Define portable agents with declarative manifests, then run them locally, inside an application, or on remote compute. Choose the model providers, infrastructure, and adapters that fit your deployment.

npm install @clearideas/agent-runtime

Agent Runtime is an independent Apache-2.0 open-source project. It offers a manifest-first alternative for teams evaluating LangChain, LangGraph, or a custom agent harness: agent definitions remain portable while the host controls models, credentials, tools, storage, compute, and telemetry.

Quick startRun your first agent with one model API key.Build agentsLearn prompt chains, routing, loops, tools, and approvals.Embed Agent RuntimeCompose the TypeScript API inside an application.Remote executionRun agents through HTTPS workers or remote compute adapters.

How it fits together

Agent manifest
Reusable definition

→

Agent run manifest
Reference and inputs

→

Execution graph
Resolved dependencies and scheduling

→

Adapters
Storage and compute

The agent manifest defines the agent. An agent run manifest references that definition and supplies invocation values. Host configuration selects model providers, credentials, and connections. Adapters connect persistence, compute, sandbox providers, and telemetry.

Agent Runtime resolves manifest variable dependencies into an execution plan. Sequential mode follows manifest order. Parallel mode schedules eligible independent prompt branches concurrently, then commits their results in manifest order; dependent, stateful, and tool-enabled work remains ordered.

Authorization policy, sandbox contracts, durable run state, and execution engines are native parts of Agent Runtime. Agent manifests may narrow host-authorized connections and tools, but cannot grant themselves credentials or broader access. Modal remote execution can apply a host-resolved direct domain list, allow only a host-provided proxy endpoint, or block networking without adding egress fields to the manifest. Docker and Modal sandbox providers ship as packages in the same project.

Choose a path

Goal	Start here
Try Agent Runtime locally	Quick start
Learn the agent patterns	Build agents
Use local or hosted models	Models and providers
Add MCP or application tools	Connections and tools
Add Agent Runtime to a service	Embed Agent Runtime
Choose packaged integrations	Adapter catalog
Run on ad hoc compute	Remote execution
Prepare a hosted deployment	Production guide

Execution model

• Agent manifests are YAML or JSON documents containing serializable values.
• Model, persistence, compute, sandbox, and telemetry integrations use adapters.
• Checkpoints record committed state and nested progress.
• Agent manifests resolve into dependency-aware execution plans.
• Parallel mode can fan out eligible prompt branches while preserving deterministic orchestration and manifest-order commits.
• Stateful steps and tool calls remain sequential.
• Runtime policy and host configuration authorize credentials, connections, tools, network access, and compute.