> ## Documentation Index > Fetch the complete documentation index at: https://docs.coverbase.com/llms.txt > Use this file to discover all available pages before exploring further. # How the agent differs from earlier AI features > Compare the Coverbase AI agent (Beta) with the Ask Coverbase AI chat, the AI inside assessments, questionnaires, intake, contracts and Radar, the Intake Agent and the MCP server: what each reads and writes, who approves, and where each is recorded. For AI agents: a documentation index is available at [https://docs.coverbase.com/llms.txt](https://docs.coverbase.com/llms.txt). This page is also available in markdown by appending .md to the URL. > [!NOTE] > > This page is part of the [User Guides](https://docs.coverbase.com/user-guides/overview.md) collection and the **Coverbase AI agent (Beta)** set. It sits beside [The Coverbase AI agent (Beta)](https://docs.coverbase.com/user-guides/coverbase-ai-agent-overview.md), [What the agent can do](https://docs.coverbase.com/user-guides/coverbase-ai-agent-capabilities.md) and [Controls, approvals and governance](https://docs.coverbase.com/user-guides/coverbase-ai-agent-controls.md). > [!NOTE] > > Coverbase AI is in Beta. Your Coverbase team turns it on for your organization. The Coverbase AI agent is the first AI in Coverbase that carries out a multi-step job across the product and writes records as you. Every earlier AI feature either answers a question, proposes a change you apply yourself, or produces one output inside one feature for a person to review. The mistake people make most often is assuming the agent replaces the earlier features. It does not. Control evaluations, questionnaire autofill, Document Insights, intake matching, clause review and Radar summaries keep running as they do today, and the agent calls the same features when its plan needs them. What is new is a single run that moves between them. ## At a glance | | Ask Coverbase AI chat | AI inside a feature | Intake Agent | Coverbase AI agent (Beta) | | - | - | - | - | - | | What starts it | You ask a question on a vendor, an assessment or a Radar signal | The feature runs: an assessment launches, a questionnaire is autofilled, a document arrives, a signal comes in | A requester messages Slack, Microsoft Teams, email or an AI assistant | You give an instruction in Coverbase AI, a chat message, or a standing agent you set up | | Reads | The record you are on and related records | The documents and answers that feature works on | The requester's message and your vendor list | Any record you can read, the page you are on, files you attach, and public web search | | Writes | Nothing until you apply a proposal from its card | Its own output: an evaluation, a draft answer, an extracted value, a summary | An intake request, as the requester | Records across the product, as you | | Shape of the work | One answer per question | One output per item | One conversation, then one intake request | A plan of many steps, carried out in order, with waits of minutes to days | | Who approves | You, by applying a proposal | A reviewer, when they review the feature's output | The Front Door reviewers, when they triage the request | You, at each checkpoint, within the organization policies your admins set | | Can it reach a vendor | No | Only through the feature's own vendor step, such as a portal you sent | No, it writes to the requester | Only when you ask, and every send stops for your approval | | Runs without you watching | No | Yes, as part of the feature | Yes, per message | Yes, in the background or as a standing agent | | Where it is recorded | The chat itself. A proposal you apply is your own change on the record | The Agent Ledger, under the agent that feature uses | The Agent Ledger, under **Intake agent**, and the Intake Agent's conversations list | The Agent Ledger, under **Coverbase AI**, plus the run's log and report | ## Ask Coverbase AI chat The **Coverbase AI** button on a vendor, an assessment or a Radar signal opens a chat about that record. It answers in one reply, from the record and its related records, and can search the public web when your organization has web research turned on. On a vendor or an assessment, the chat can propose a small set of changes, such as updating the vendor's profile, adding an owner, raising a finding, marking a control as an issue or not, or drafting a follow-up. Each proposal shows as a card, and nothing changes until you apply it. A drafted follow-up never reaches the vendor until someone sends it. The chat keeps working when Coverbase AI is on. Use the chat to ask about the record in front of you. Use Coverbase AI to get a job done. | Question | Ask Coverbase AI chat | Coverbase AI agent | | - | - | - | | "What is still open on this assessment?" | Answers from the assessment | Answers from the assessment, with a link for every record it cites, and writes nothing | | "Mark control 4.2 as not an issue" | Proposes the change on a card for you to apply | Proposes it at a checkpoint, or applies it when you set the run to **Don't stop to ask** | | "Review this assessment" | Not available: the chat answers one question at a time | Goes through every control that needs a person and proposes an outcome for each | | "Onboard Figma" | Not available | Runs intake, screening, the inherent risk questionnaire, the assessment and monitoring | ## AI inside a feature Coverbase has used AI inside individual features for some time. Each one produces one kind of output, and a person reviews it in that feature. The Agent Ledger records them under these agents: | Feature | What its AI does | Ledger agent | | - | - | - | | Assessments ([Assessment Copilot](https://docs.coverbase.com/products/assessment-copilot.md)) | Evaluates each control against the evidence and drafts follow-ups for gaps | Evidence agent | | [Zero Touch Assessments](https://docs.coverbase.com/products/zero-touch-assessments.md) | Evaluates a vendor from public evidence, with no outreach | Evidence agent | | [Claim Check](https://docs.coverbase.com/user-guides/claim-check.md) | Checks a vendor's answers against its SOC report | Evidence agent | | Questionnaire autofill | Drafts answers to a questionnaire from documents | Drafting agent | | [Autonomous Intake](https://docs.coverbase.com/products/autonomous-intake.md) | Matches the vendor, plans questions and suggests follow-ups during intake | Intake agent | | [Contract Guardian](https://docs.coverbase.com/products/contract-guardian.md) | Reviews contract clauses against your standards | Contract agent | | Risk narratives | Drafts a risk report narrative | Risk agent | | [Radar](https://docs.coverbase.com/products/supplier-radar.md) | Summarizes a Radar signal | Radar agent | | [Monitoring plans](https://docs.coverbase.com/user-guides/monitoring-plans.md) | Moves monitoring activities when a signal fires, if your policy allows | Monitoring agent | | [Document Insights](https://docs.coverbase.com/products/document-insights.md) | Extracts the fields you define from each document, with a quote, a page and a confidence | Not a ledger agent: each value carries its own source | These features decide nothing on their own beyond their output. A control evaluation sets an outcome a reviewer can overturn. An autofilled answer is a draft. The agent policies under **What Agents May Do** on the Agent Ledger page govern the two that can act on their own today. See [Choose what agents may do](https://docs.coverbase.com/user-guides/agent-ledger.md#step-3-choose-what-agents-may-do). Coverbase AI calls these features rather than replacing them. When it launches an assessment, the assessment's controls are evaluated as they would be for an assessment you launched. When it starts autofill on a questionnaire submission, autofill drafts the answers as it does when you start it. ## Intake Agent The [Intake Agent](https://docs.coverbase.com/user-guides/intake-agent.md) is for requesters. It reads a request in Slack, Microsoft Teams, email or an AI assistant, fills in what the message says, asks for what is missing, and opens an intake request that the [Front Door](https://docs.coverbase.com/user-guides/front-door-triage.md) triages. It acts for the requester and stops at the intake request. Coverbase AI is for your risk team. It can open an intake request too, through the same intake portal, and then carry on: screening, the inherent risk questionnaire, submitting the request (a checkpoint), the inherent risk review, the assessment and monitoring. On a submitted request it can read the Front Door review, send the request back to the requester with questions, or deny it with a reason. Each of those always asks first. In Slack and Microsoft Teams, a direct message that starts with `Coverbase AI:` starts a Coverbase AI run instead of an intake request, when your organization has Coverbase AI and your role can run it. Every other message stays a request to the Intake Agent. ## Your AI assistant through MCP The [Coverbase MCP server](https://docs.coverbase.com/mcp/overview.md) lets your own AI assistant read and change Coverbase records with your permissions. Every write it sends needs your confirmation first. It can also start, answer, steer and stop Coverbase AI runs (`mutate(entity="agent_run", ...)`). A run started from MCP is an ordinary background run with the same checkpoints, policies and ledger rows as one started in the app. See the [MCP tool reference](https://docs.coverbase.com/mcp/tool-reference.md#coverbase-ai-and-agents). ## What is the same * **Models and hosting.** Coverbase AI and the earlier features run on Anthropic Claude models through AWS Bedrock in Coverbase's AWS account. See [AI governance](https://docs.coverbase.com/security/ai-governance.md). * **No training on your data.** Customer data sent to the models is not used to train them. * **Redaction.** When **Redact personal data before model calls** is on, it applies to Coverbase AI's model calls as it does to every other model call. * **Tenant isolation.** All of it runs inside your organization's boundary. ## Troubleshooting | Problem | What to do | | - | - | | The Coverbase AI button on a vendor opens a chat, not a run | That is the Ask Coverbase AI chat. Open **Coverbase AI** in the left navigation or the Coverbase AI button in the top bar to start a run. | | You expected a control evaluation to change after a run | Coverbase AI changes an evaluation only at a checkpoint you approve, or under **Don't stop to ask**. Check the run's report and the ledger. | | A requester's Slack message started a run instead of an intake request | The message started with `Coverbase AI:`. Ask requesters to write without that prefix. | | You cannot find a chat's applied proposal in the Agent Ledger | A proposal you apply from the chat is your own change. Look in the record's history. | ## Related > [card] [The Coverbase AI agent (Beta)](https://docs.coverbase.com/user-guides/coverbase-ai-agent-overview.md) > > What the agent is and how a run works. > [card] [Controls, approvals and governance](https://docs.coverbase.com/user-guides/coverbase-ai-agent-controls.md) > > How the agent's changes are controlled and recorded. > [card] [Agent Ledger](https://docs.coverbase.com/user-guides/agent-ledger.md) > > Every action an agent took, and who decided. This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.