> ## Documentation Index
> Fetch the complete documentation index at: https://docs.usenash.com/llms.txt
> Use this file to discover all available pages before exploring further.
# Daily Brief
> A morning read of what your agents found — curated insights with evidence you can verify, and the decisions waiting on you.
The Daily Brief is the page to open first. It gathers what your agents noticed across their runs into a short list of insights worth a person's attention, each with the evidence behind it, and keeps the confirmations waiting on you in one place. It sits under **Daily Brief** in the Nash Agent sidebar.
## When to use this
* You run several agents and don't want to open each one's sessions to find out whether anything happened overnight.
* An agent in auto-execute mode has paused on a confirmation and you want the queue of those decisions, not a notification each time.
* You want to check an agent's claim before acting on it — the brief links every observation to the run that produced it.
For one delivery or route you care about, use the [entity monitor](/nash-agent/entity-monitor); its watch session is the running log. For everything an agent did, open the agent's sessions. The brief is deliberately short.
## How it works
Every custom agent records what it notices as it runs. Throughout the day Nash reads those signals across your organization and curates them into **Agent insights**: a pattern across deliveries rather than one agent's single observation, with what Nash found and what it recommends. The brief opens with a one-line summary — how many agents ran, how many times, and how many new or changed insights are waiting — and lists up to twenty insights, unread first.
Each insight carries:
| Part | What it holds |
| ------------------------------------------------ | --------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- |
| **What Nash found** and **Nash AI · Recommends** | The pattern, and the next action Nash suggests. |
| **Contributing agents** and **Cited posts** | Which agents' observations support it, and how many. |
| **First evidence** and **Latest evidence** | When the pattern was first seen and when it was last confirmed. |
| **Verify in Nash** | A link per cited observation to the exact agent session and message that produced it, with the recorded time. |
| **Ask Nash** | Opens a chat with the insight attached, so you can investigate and get a safe next step. The insight is handed to the chat as evidence to examine, not as instructions to follow. |
References an agent wrote into its findings — a route, an order, a job — are shown, but only become links once Nash has confirmed the ID exists. What you can't verify, you can't click.
### Read, updated, dismissed
The brief is yours. An insight is **New** until you open it. If new evidence lands on an insight you've already read, it comes back as **Updated** rather than staying quiet. **Dismiss** clears an insight from your own brief; a colleague's brief is unaffected. The **Daily Brief** entry in the sidebar counts what you haven't read.
### Awaiting your decision
Confirmations sit apart from insights, under **Awaiting your decision**. When an agent in auto-execute mode reaches an action your [guardrails](/nash-agent/guardrails) hold for approval, the card appears here with a **Review** action, and nothing happens until you answer.
### What the brief is for
The brief carries findings about your operation: a pattern across deliveries, an outside event affecting them, a data problem in the orders you send. A finding that is really about Nash itself, or about how one of your agents is configured, is routed to Nash's team rather than to your brief, so what you read is what you can act on.
### On each agent's page
A custom agent's own page shows its **Fresh signals**: **Recent observations** it recorded in the last 24 hours, and the **Org insights this agent contributed to** — the brief's patterns that its observations support. It's the quickest way to see whether an agent is noticing anything at all.
## FAQ
New evidence arrived on it. Rather than leave a read insight looking settled while agents keep confirming the pattern, Nash marks it **Updated** and moves it back to the front.
No. Read and dismissed states are per person. Dismissing tidies your own brief and nobody else's.
Anyone whose role can read Nash Agent. Confirmations still require the permission to act on the underlying action.
The brief is curated, and capped at twenty insights at a time. Open an agent's page for its own recent observations, or its sessions for the full record.
## Related
The agents whose runs feed the brief.
Keep watch on one delivery, order, or route until it's done.
Which actions pause for your approval.
How runs start, and what happens when they finish.