Agent OS FAQs
Agent OS is Contentstack’s adaptive intelligence framework that unifies Agents, Automations, and Polaris to deliver intelligent, context-aware, and governed digital operations. It blends deterministic workflows with LLM-powered reasoning to help enterprises operate faster and more intelligently.
Agent OS is made up of four main components:
Agents: The intelligent reasoning layer that understands context, makes decisions, and performs actions using instructions, tools, and brand governance.
Automations: The workflow engine that runs event-driven and scheduled tasks with reliability and governance.
Polaris: The in-app conversational assistant that provides contextual guidance and interacts with agents and automations.
No, it's designed for business and technical teams with a no-code interface.
No, it enhances them. Agents and Automations work together seamlessly.
Absolutely. It connects via connectors and MCP for external systems like Slack, Jira, Notion, and others.
An Agent is an intelligent, AI-powered system within Contentstack Agent OS that can understand context, reason, make decisions, and take actions on behalf of users. Unlike traditional automations that follow fixed, rule-based steps, agents use a combination of:
Model: LLM for reasoning and language understanding
Instructions: Persona, guidelines, goals, boundaries
Context: Situational data and enterprise knowledge for adaptive decisionsTools/Abilities: Actions, automations, and other agents they can call
No. Agents can be fully created and configured using a no-code interface.
Agents run within governed enterprise controls, use secure authentication, and follow brand and compliance rules.
You can review execution logs, adjust instructions or abilities, and refine behavior. Brand Kit and Knowledge Vault minimize errors.
Polaris is Contentstack’s in-app conversational assistant that provides intelligent, context-aware help across the CMS. It allows users to interact with the platform through natural language, asking questions, generating content, running tasks, or triggering agents and automations.
Polaris also uses brand context and system knowledge to deliver accurate, on-brand responses.
Yes. Polaris is accessible across the platform.
By default, only users with the Admin or Owner role in Administration have access to Polaris. Members do not see the Polaris icon unless an admin explicitly grants them access via a Custom Role.
Your account likely has a Member role, which does not include Polaris access by default. Contact your organization admin and ask them to assign you a Custom Role with Polaris Read access. If your admin has already updated your role, try refreshing your browser, the Polaris icon should appear after a refresh.
- Go to Administration > Roles
- Click Create Custom Role and give it a name (e.g., "Polaris Viewer")
- Enable Read access under the Polaris section and saveGo to the user's profile, assign the new Custom Role, and save
The user will see the Polaris icon after refreshing their browser.
No. A simple browser refresh is enough. Access changes take effect immediately, no re-login required.
Yes. Once the Custom Role with Polaris Read access is created, you can assign it to as many users as needed, no need to create a separate role for each user.
- Confirm the Custom Role has Polaris Read access enabled, not just created with empty permissions
- Confirm the Custom Role is actually assigned to the correct user
- Ask the user to do a hard refresh (Ctrl+Shift+R on Windows/Linux, Cmd+Shift+R on Mac)
- If the issue persists, ask the user to log out and log back inIf still unresolved, contact Contentstack Support
The current documented flow covers Read access. If you need to explore additional permission levels, check the available options when creating your Custom Role, or reach out to Contentstack Support.
Automations help you to perform repetitive tasks without human intervention. You can set up an automation using various third party applications. For example, you can notify your team via an email whenever an entry is created in your stack.
To set up an automation, you need to configure your trigger and action connectors. To learn more, refer to the Create Automation step in the Get Started with Automations document.
Connectors are a combination of Triggers and Actions events. When a trigger event is executed, the action connector performs the defined steps. Refer to Connectors to get started.
Trigger connectors are conditions that help set off an automation whenever the selected event is executed. Action connectors is the action performed when a trigger event is executed. For example, an automation is set up to send a slack message (action event) whenever an entry is created (trigger event) in Contentstack.
You can configure only one trigger event in a particular automation. You can configure multiple actions for a single automation. For example, you can set up an automation to perform end to end translation when an entry is created/updated in Contentstack. For this, you need to configure different action connectors to carry out the translation process. To learn more, refer to our End to End Translation using Smartling example.
You can view the logs for your automation in the Execution Log section. This is useful for checking the status of your automations. To learn more, refer to our documentation on Execution Log.
You can also monitor the acitvities performed in a particular project via the Audit Log.
You can view all the connected apps for your project in the Connected Apps section. You can check the authentication status of an app or re-authorize an app. To learn more, refer to our documentation on Connected Apps.
Once you create an Automation, the list of existing automations appears in the Automations listing page. In the Actions column, click the three ellipses icon, and then click the Edit icon that you see for editing the automation. Refer to our documentation on Edit Automation Details.
On the Automations listing page, click the three ellipses icon, and then click the Delete icon that you see for deleting the automation. Refer to our documentation on Delete Automation.
In the automation, click the edit icon visible on the action header to rename the step. Add a suitable name for the step and save it. Refer to our documentation on Rename Step.
To re-authorize a connected app, follow the steps given below:
- In the Connected Apps section, select the app that you want to re-authorize.
- Click over the connection and click the Reauthorize icon.
- Change necessary permissions and click the Authorize button.