> ## Documentation Index > Fetch the complete documentation index at: https://docs.bland.ai/llms.txt > Use this file to discover all available pages before exploring further. # Norm > Bland's AI assistant for building, editing, testing, and debugging voice agents from a prompt. ## Introduction Norm is the AI assistant built into Bland. You describe what you want in plain English, and Norm builds, edits, tests, and debugs your agents for you. Agents, tools, analytics, and triage all work from one conversation. Norm works the way an experienced Bland engineer would: it inspects the live state of your account before making changes, works in development so production keeps running, tests against real call data, and only ships changes when you're ready. > [card] Build and edit agents > > Describe an agent, a node, a route, or a tool change. Norm makes the change in development, where you can review it before it goes live. > [card] Test and debug > > Run simulated conversations and trace failed calls back to a root cause. > [card] Query analytics > > Ask in plain English: "how many calls failed yesterday?", "transfer rate by agent this week" and get tables, charts, or a single number. > [card] Fix issues from Triage > > Flag a bad call, and Norm reproduces it, finds the root cause, applies the fix, and verifies it against the original call. *** ## Where to find Norm > [card] Norm chat > > The chat surface in the Bland dashboard. It opens full page on the Overview and in the side panel everywhere else. Use it for any task across your org. > [card] [iMessage](https://docs.bland.ai/tutorials/messaging/imessage.md) > > Text Norm at **+1 (321) 424-0172** to build, test, and monitor agents from your phone. See [Norm over iMessage](https://docs.bland.ai/tutorials/messaging/imessage.md#norm-over-imessage). > [card] [Claude Code](https://docs.bland.ai/integrations/mcp/norm.md) > > Install the [Norm plugin](https://docs.bland.ai/integrations/mcp/norm.md) to build, test, and ship agents without leaving your terminal. ![The Norm chat interface in the Bland dashboard](https://mintcdn.com/blandai/cn3mkn0afMFA51cl/tutorials/tutorials-assets/Norm_dashboard.png?fit=max\&auto=format\&n=cn3mkn0afMFA51cl\&q=85\&s=5fa18aca60f70c23aad6c770564d46c1) ### Norm on the Overview The all-agents Overview in the [Bland dashboard](https://app.bland.ai/dashboard) is the home for your Norm conversations. * **Conversations table.** Your own Norm conversations appear under the address book, newest first, with a **Last active** column. The table shows 10 rows. Select **Show all** to see older conversations. * **Full-page chat.** Select a row, ask a question in the Overview composer, or select **New chat** to open the conversation full page. A recent-chats rail on the left lets you switch conversations, and you can hide it. Select **Exit** at the top right to return to the Overview. * **Row menu.** Open a row's menu to **Copy conversation ID** or **Rename** the conversation. Renaming doesn't change the conversation's position in the list. On the Overview, Norm opens full page instead of in the side panel. On every other page, Norm opens in the side panel as before. If you leave the Overview with a conversation open, that conversation continues in the side panel. To move a side panel conversation onto the full page, select the thread button in the side panel header. *** ## What Norm can do ### Build and edit agents Norm can create a new agent from a description or edit an existing one, including its prompts, pathways, tools, knowledge, and settings. Common requests: * *"Build me a scheduling agent that books appointments on Cal.com."* * *"Add a node after intake that verifies the caller's policy number against our CRM."* * *"Change the transfer fallback on this agent to go to billing instead of sales."* Norm edits each part of an agent through the right interface rather than hand-writing raw config. ### Test and debug Norm has direct access to your call logs and transcripts. When you point it at a problem, it gathers the evidence before suggesting a fix. Common requests: * *"Why did call `abc-123` fail to transfer?"* * *"This agent is misrouting refund requests to the cancellation flow. Find out why."* To check a fix on the exact turn that went wrong, use [Testbed](https://docs.bland.ai/agents/testbed.md). It re-runs one turn of a real conversation with your changes, without redialing anyone. ### Query analytics Ask any question about your call data in plain English. Norm picks the right aggregation, time range, and visualization for the question. Common requests: * *"How many calls did I get in the last 10 minutes?"* * *"Transfer rate by agent over the last 30 days, as a bar chart."* * *"Compare average call duration this week vs. last week."* * *"Top 5 disconnect reasons for failed calls yesterday."* ### Fix issues from Triage Flag a call as a Triage issue and hit **Norm Fix**. Norm pulls in the failing call and its agent, reproduces the failure, applies a fix, and verifies it against the original call. Every change Norm makes is reviewable on the issue, so you can see exactly what it touched before the fix reaches production. ### Run skills Skills are reusable workflows that scope Norm to a specific task, a built-in skill like `/systematic-debugging`, or a custom skill your team writes. Type `/` in Norm chat to open the skill menu. Custom skills are scoped to your organization and let you encode anything your team does repeatedly: brand-voice enforcement, common agent scaffolds, or your standard QA scenarios for every new agent. *** ## How Norm keeps production safe Norm never edits what live calls run. A call runs the version pinned to **production**, and every change moves through these stages: **1. Edit in development** Norm makes changes in the agent's **development** environment, the only one you edit. Production keeps serving its pinned version while you and Norm work. **2. Save a version** Once the change passes validation and any tests you run, Norm saves it as a new version. Saving a version does not change live traffic. **3. Publish and promote** To put the change on live calls, the version is published to **staging** and then promoted to **production**. An agent's first publish goes straight to production. If evaluations are configured on staging, the required ones must pass first. See [Environments](https://docs.bland.ai/agents/environments.md). > [!NOTE] > > Approval to save a version is not approval to move live traffic. Production is untouched until the version is promoted. *** ## Tips for working with Norm * **Anchor on a real call when possible.** If you're debugging, paste the call ID. Norm will pull the full call and reproduce the failure before suggesting a fix. * **Be specific about scope.** Norm follows surgical-edit rules: it changes what you asked for and nothing else. If you want broader cleanup, say so. * **Ask Norm to explain before it edits.** "Walk me through what you'd change before doing it" is a valid prompt. * **Use skills for repeatable work.** If you find yourself giving Norm the same instructions repeatedly, turn them into a custom skill. *** ## Related docs > [card] [Agent builder](https://docs.bland.ai/agents/agent-builder.md) > > The prompts, pathways, tools, and settings Norm edits when you ask for agent changes. > [card] [Environments](https://docs.bland.ai/agents/environments.md) > > How a saved version is published to staging and promoted to production. > [card] [Testbed](https://docs.bland.ai/agents/testbed.md) > > Re-run one turn of a real conversation to check a fix. > [card] [Evaluations](https://docs.bland.ai/agents/evaluations.md) > > The judges and test cases that score your agent before it ships. > [card] [Tools](https://docs.bland.ai/tutorials/tools/overview.md) > > The tool library Norm authors and attaches to your agents. > [card] [iMessage](https://docs.bland.ai/tutorials/messaging/imessage.md) > > Talk to Norm over iMessage from your phone. This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.