Categories
Self-Hosted AI Agents AI Infrastructure Agent Orchestration Multi-Agent Systems Control Plane Agentic AI Workflow ObservabilityStars
Forks
Watchers
Developer links
AgentField
AgentField is the production backend that makes an AI agent as callable as any other microservice. A Python, Go, or TypeScript function decorated as a reasoner or skill registers with the control plane and instantly becomes a REST endpoint, and agents call each other through the same plane, so every hop is routed and traced. The dashboard summarizes runs, success rate, online agents, and a run timeline, and a Needs attention list surfaces failed runs with a one-click jump into the DAG. The Runs page filters and searches every execution, and a run opens into a step-by-step trace with timings, inputs, and outputs, or a graph of the call tree across agents. Agent Nodes lists each registered service with version, endpoints, and process logs. The Playground executes any reasoner from a form and shows the result and recent runs. Fan-outs run async with webhooks, server-sent events, bounded queues, and backpressure, and a single run may continue for hours. Human approval pauses, canary traffic weights, Prometheus metrics, access policies, and signed provenance records cover governance. Running on a dedicated VPS on RepoCloud with guaranteed CPU, RAM, and SSD, full root SSH access, and a browser serial console. Apache-2.0 licensed.
Benefits
- Agents Callable Like APIs
- Every reasoner and skill becomes an HTTP endpoint, so frontends, backends, cron jobs, and other agents can call your agent logic without custom queues or glue code.
- See Every Branch Of A Run
- Each execution records a trace and call graph across agents with timings, inputs, and outputs, so a failure in a deep fan-out is found in seconds.
- Reliable Long-Running Work
- Async execution with webhooks, retries, bounded queues, and backpressure keeps large fan-outs from overwhelming downstream agents, and runs can last hours without timeouts.
- Governance Built In
- Agents receive cryptographic identities, tag-based access policies, and signed execution records, giving auditors verifiable proof of which agent did what.
Features
- Run Explorer
- Search and filter every execution, then open a run to inspect its step trace, logs, inputs, outputs, and replay or restart it from the same screen.
- Workflow Graph View
- Switch a run from trace to graph to see a DAG of reasoners and skills across agent nodes, with per-step durations on each edge.
- Agent Node Registry
- List every registered agent with version, health, reasoners, skills, endpoints, auth tags, and process logs, and restart a node from the table.
- Reasoner Playground
- Pick any reasoner or skill, edit its JSON input, execute it, copy the matching cURL command, and review recent runs of the same function.
- Access And Provenance
- Manage agent tags and policies, publish discovery catalogs, and export verifiable provenance for an execution when compliance teams ask for evidence.