> ## Documentation Index
> Fetch the complete documentation index at: https://insight-flow.potatodog.cc/llms.txt
> Use this file to discover all available pages before exploring further.

# Monitors

> The monitor model, configuration fields, and runtime workflow

A monitor is the core scheduling unit in Insight Flow. It combines sources, AI routing, report type, and destinations into a repeatable run configuration.

## What a Monitor Contains

* Name
* `time_period`
* `report_type`
* `source_ids`
* `ai_routing`
* `destination_ids`
* `window_hours`
* `enabled`

## Key Capabilities

* Create and update monitors
* Trigger runs manually
* Inspect runs, logs, and events
* Cancel a running job

## How to Use It

1. Configure sources, providers, and destinations first
2. Create a monitor and choose its sources and destinations
3. Set providers for `filter`, `keywords`, `global_summary`, and `report`
4. Save and run it, then inspect runs and logs

## Report Type Relationship

* `daily`: ongoing monitoring
* `weekly`: periodic summaries
* `research`: longer-form research output

## Related API

* List and create: `GET /api/v1/monitors`, `POST /api/v1/monitors`
* Default AI routing: `GET /api/v1/monitors/ai-routing/defaults`
* Run: `POST /api/v1/monitors/{monitor_id}/run`
* Logs: `GET /api/v1/monitors/{monitor_id}/logs`
* Runs: `GET /api/v1/monitors/{monitor_id}/runs`
* Cancel: `POST /api/v1/monitors/{monitor_id}/runs/{run_id}/cancel`
* Events: `GET /api/v1/monitors/{monitor_id}/runs/{run_id}/events`
* Update and delete: `PATCH /api/v1/monitors/{monitor_id}`, `DELETE /api/v1/monitors/{monitor_id}`

## Recommended Workflow

* Validate a small source set first
* Confirm destination output before enabling long-form monitors
* If a run stalls, inspect runs first, then events and logs