How to Run a Long-Running AI Agent with Navin

Navin can keep agent work alive across turns through sustained goals, persistent chat sessions, scheduled automations, and heartbeat checks - while the desktop app stays open.

What you will use

a working local agent in the Navin app
a persistent chat session
a long-running goal or automation
Navin left running for background delivery

When to use this

Use this when the task is not a one-shot answer: project work, recurring checks, scheduled summaries, file maintenance, multi-step research, or quiet watch loops.

Setup

Open Navin and confirm Settings → Providers / Models.
Open the project and chat where work should continue.
Leave Navin running while you need goals or schedules to continue.

Minimal working example

From chat, start a sustained goal:

/goal Review this project, identify missing tests, and propose the smallest next fix.

For scheduled or recurring runs, ask from the same chat so Navin links the automation to the correct session and project. Details: ../automations.md.

For quiet checks that should only notify when something matters, edit HEARTBEAT.md in the project and leave Navin open.

Production notes

Keep Navin open for chat apps, automations, and heartbeat.
Use stable chat sessions for work that should preserve context.
Keep goals bounded and explicit about done-ness.
Review Automations in the app before relying on a schedule.

Security notes

Treat long-running goals as delegated work with real tool access.
Restrict projects and shell execution before scheduling unattended tasks.
Keep chat channel access narrow so unknown users cannot create goals or automations.

Troubleshooting

Goal appears stuck: inspect the active chat session and pause/resume from the UI when available.
Automation does not run: confirm it is linked to a chat, enabled, and that Navin is still open.
Heartbeat silent: check HEARTBEAT.md and heartbeat settings; silence is often intentional when nothing actionable appeared.