LightfieldAI-native CRM that builds itself and does work for you
Promoted
Launching today
Cronus
Durable background execution for AI agents
0 followers
Durable background execution for AI agents
0 followers
Cronus is local-first durable execution infrastructure for AI agents. Run background jobs, DAG workflows and schedules with checkpoints, retries, leases, recovery, dead letters and approval gates. Work can survive agent, worker and process restarts instead of disappearing with the session. Cronus runs locally on Windows and Linux and connects to AI clients through MCP over stdio, with local REST and gRPC interfaces.
Hey Product Hunt 👋
I built Cronus as part of a broader idea I've been working on with Titans: foundational capabilities that AI agents repeatedly need shouldn't have to be redesigned and rebuilt inside every new project.
They should exist as reusable infrastructure that new agents and applications can simply consume.
Cronus handles one of those recurring problems: durable execution.
An agent may start work that takes minutes or hours, requires retries, depends on another task, waits for human approval, or needs to continue after the original agent session or process has disappeared.
Cronus lets that work move outside the lifetime of the agent that initiated it.
It provides:
• Background jobs
• DAG workflows
• Scheduling
• Checkpoints and retries
• Leases and recovery
• Dead-letter handling
• Approval gates
• MCP over stdio
• Local REST and gRPC interfaces
Cronus is local-first and agent-first, currently runs on Windows and Linux, has no telemetry, and is distributed through signed releases.
We're also releasing Atlas, the other first system in the Titans ecosystem. Atlas provides persistent memory, state, knowledge and evidence across agents and sessions.
Atlas remembers. Cronus runs.
Both can operate independently or together, and the intention is that existing applications, new agent systems and other parts of Titans can consume them rather than rebuilding these capabilities themselves.
Other tools are already being implemented as part of the ecosystem, following the same principle: identify infrastructure problems that repeatedly appear in agentic systems, solve them once as reusable capabilities, and make them available to future systems.
I'd especially value feedback from people building agent systems:
• Which infrastructure problems do you find yourself rebuilding across projects?
• How are you currently handling work that needs to survive the agent or session that started it?
• What would you need to see before using Cronus as part of your own agent infrastructure?
Critical technical feedback is very welcome.