> ## Documentation Index > Fetch the complete documentation index at: https://docs.jengolabs.com/llms.txt > Use this file to discover all available pages before exploring further. # Agent persona > How the persona works and what each tab does ## What is the persona? The **persona** is where you configure how an agent behaves: what it says, what it can read, which rules it follows, and which workflows it can run. You open it from an agent (e.g. **Persona** button or the agent’s settings). The persona is split into **tabs**; each tab controls one area. Agent persona — tabs (General, Knowledge base, Behaviours, Workflows, Voices, Actions) *** ## General **What it does:** Defines the agent’s personality and instructions — how it should speak and act in every conversation. **What you set:** * **Personality / instructions** — Free-form text that tells the agent its role, tone (e.g. friendly, professional), and any hard rules (e.g. never share internal links, always confirm before creating a work order). The agent uses this as its base behavior. * If you started from a **template**, this tab is often pre-filled; you can edit it to match your brand and policies. **Tip:** Be specific about tone and boundaries so the agent stays consistent and safe. *** ## Knowledge base **What it does:** Chooses which content from your **Knowledge base** this agent is allowed to use when answering (folders, documents, links). **What you set:** * A list of your organization’s knowledge base items (folders and files). You **select** which ones this agent can use. Only selected items are used to answer questions. * The agent searches this content and uses it in its replies so answers stay accurate and on-brand. **Tip:** Give each agent only the knowledge that’s relevant (e.g. tenant support gets tenant FAQs; internal ops might get different folders). See [Knowledge base](/agents/knowledge-base) for adding and organizing content. *** ## Behaviours **What it does:** Turns on or off **predefined behaviours** (e.g. “always be polite”, “escalate if unsure”). Behaviours are created by your admin and shared across agents. **What you set:** * A list of available behaviours. You **toggle each one on or off** for this agent. Only behaviours you turn on apply. * Behaviours add rules or style without you writing long instructions in General. **Tip:** Use behaviours for tone, escalation, and safety so you don’t duplicate the same rules in every agent. See [Behaviour](/agents/behaviour) for more. *** ## Workflows **What it does:** Chooses which **workflows** this agent is allowed to run (e.g. create work order, send notification, add to leads). **What you set:** * A list of workflows your organization has. You **select** which ones this agent can trigger. When a user’s message implies an action (e.g. “I need to report a leak”), the agent can run one of these workflows if you enabled it. * Restrict to the workflows that make sense for this agent so it doesn’t run the wrong process. **Tip:** Test in the preview chat with phrases that should trigger a workflow and confirm the right one runs. *** ## Voices **What it does:** (Only for **voice** or **phone** agents.) Sets the voice and language used when the agent speaks on a call. **What you set:** * **Voice** — Which synthetic voice to use (e.g. from a list of available voices). * **Language** — Language for speech (e.g. English, French). This tab is shown only for agent types that support voice (e.g. Phone agent, Voice agent). *** ## Actions **What it does:** (Only for **voice** agents.) Configures what the agent can do during a call (e.g. transfer, book appointment, look up information). **What you set:** * Which **actions** are available during a call (e.g. transfer to human, create task, check calendar). You enable or configure the ones that fit this agent. * The agent can use these actions when the user asks for something (e.g. “Transfer me to support”, “Schedule a visit”). This tab is shown only for voice/phone agents. The exact list depends on your plan and integrations. *** ## Summary | Tab | Purpose | | ------------------ | -------------------------------------------------------------- | | **General** | Personality and instructions (tone, rules). | | **Knowledge base** | Which folders/documents/links the agent can use to answer. | | **Behaviours** | Predefined behaviours on/off (tone, escalation, safety). | | **Workflows** | Which workflows the agent can run. | | **Voices** | Voice and language (voice/phone agents only). | | **Actions** | What the agent can do during a call (voice/phone agents only). | Save after changing a tab if the app asks you to. Then test the agent in **Preview** or **Test** chat (and for voice agents, with a test call) to confirm everything works as expected. ## Related * [How to create an agent](/agents/create-agent) — Where the persona fits in the flow. * [Behaviour](/agents/behaviour) — Managing behaviours. * [Knowledge base](/agents/knowledge-base) — Adding and linking content.