How to use agent mode in WRITER Agent

What's included in this article?

How to use agent mode

Agent mode takes your sessions to a new level, using information to complete tasks.

Instead of asking it to tell you how to make a presentation, you can ask it to make the presentation for you. It will find the information, structure the slides, write the content, and deliver a finished file.

From goal to finished product

The most powerful aspect of agent mode is the ability to work autonomously toward a goal. You don't need to know every step required to get something done.

You say: "Analyze the attached sales data and create a report on our top five performing products this quarter."

WRITER Agent does this: Opens the file, analyzes the data, identifies the top products, writes a summary of the findings, formats it into a clean report, and gives the final document to you.

You provide the destination, and WRITER Agent figures out the best way to get there.

To get started select the Chat/Agent toggle within the input field to active agent mode. Select the + icon to add source files, connect to external tools, select from playbooks, or enable presentation mode. WRITER Agent can handle just about any file type - if it isn’t able to handle a type of file, it has the ability to convert it or process it in a way it can use! ⚡

You can also select the microphone icon to enable voice-to-text and speak directly to WRITER.

Below the input field, we offer some suggested use cases as well as some suggested prompt completions based on what you've entered. Select one to load the prompt into WRITER Agent and see the full range of its capabilities!

Once you’ve submitted a prompt, a new window will appear on the right side. In the center pane, you’ll see WRITER Agent keep a running explanation of everything its doing, and why its doing it. On the right side, you can see a status update for each step being conducted by WRITER’s computer. You can use the scroll bar at the bottom of the panel to view previous steps in the process.

While WRITER Agent works autonomously to complete the given workflow you will have insight into what is happening within each step. The left side will show an explanation of the steps the agent works through while the Computer tab on the right displays details such as sources from the web, code being written, images being generated, etc.

Note: You can interrupt WRITER Agent at any time. If you see a fault in its logic, or simply want to change course from your original action plan, just send a followup message in the central panel.

When WRITER Agent has produced the final output(s), you’ll see them displayed on the right side within the Deliverables tab. You can select the files to expand them, download them, open them in a new window, etc.

Note: You can navigate away from WRITER Agent without interrupting your workflow. It will continue to run as long as it needs to complete its assignment.

How does agent mode work?

Working with agent mode is a simple, transparent process. You’re always in control and can see exactly what it’s doing.

Step 1: Make a request.

Begin with a clear, goal-oriented request. The more specific you are about what you want the final product to be, the better. For example, instead of “Tell me about our sales,” a better request would be, “Create a 10-slide presentation summarizing our Q3 sales performance, using the data in sales_q3.csv.”

Step 2: WRITER Agent creates a plan

Before it does anything, WRITER Agent will outline its plan to achieve your goal. It will show you a checklist of the steps it intends to take. This gives you a chance to review its approach and ensure it’s on the right track before the work begins.

Step 3: WRITER Agent gets to work

Once it has a plan, WRITER Agent starts executing it one step at a time. It has a full toolkit at its disposal, allowing it to perform tasks like:

  • Browsing the web for information
  • Reading and writing files (like documents, spreadsheets, and code)
  • Analyzing data to find insights
  • Creating new documents, presentations, and spreadsheets
  • Writing and running code to build applications or perform complex calculations

Step 4: It delivers the result

When all the steps in its plan are complete, WRITER Agent will present you with the finished product. This might be a Word document, a PowerPoint presentation, an Excel spreadsheet, or even a zip file containing a fully functional web page. You can then review the work and ask for revisions if needed.

Generating presentations

In agent mode, you can create presentation deliverables downloadable as .pptx or PDF files. To specify that you want to create a presentation, select the Presentation option from the + menu. Specify the specific fonts and colors you’d like the presentation to use. You may also consider including slide templates or style guides (via file upload or connector) for reference, along with specific URLs of logos you’d like included in the presentation (e.g., a customer’s logo).

Using Agent mode with other tools

Connectors (BETA) make it easy to bring the context you need straight into WRITER Agent so you can focus on your task at hand, not manual steps or tool management. By connecting WRITER Agent to other tools, your team can execute complex workflows to get work done efficiently. Log a meeting in Salesforce, pull sprint data from Jira, send messages to Slack or Teams, and more! Here’s how to connect other tools to WRITER Agent

When in agent mode select the + button from the bottom of the WRITER Agent input field and then select Connectors to open the menu. Within the menu you can search through the configured connectors available to you and select the ones you'd like to use from the list to activate them in your session. Select Manage connectors to launch the connector configuration panel.

Note: Connectors require admin configuration to enable use within your WRITER org. To learn how to configure connectors click here.

Once the Connector library is launched, you can search through available connectors and learn more about what you can do by selecting the three dot icon to the right and then selecting View details. The list will also display which connectors require you to login in order to activate the connection.

Within the connector details page you can see a description of what you can accomplish by connecting the application. Connectors are able to take the following actions:

  • Read- allows the agent to retrieve information from the application
  • Write- allows the agent to take action within the application
  • Delete- allows the agent to delete items within the application

Select Connect in the top right if you’d like to enable Action Agent to use the connector.

Once the connection is completed you can manage the tools WRITER Agent can use within the application. Select the dial icon to the right of each tool to enable or disable the specific actions you’d like WRITER to be able to perform.

To activate connectors within a WRITER Agent session, select the + icon from the input field and then select the connector(s) from the dropdown menu. WRITER can also auto infer what connectors to use based on the query. You only need to use the connectors menu in the + session if you want to guarantee it uses it.

Creating playbooks

If there’s a prompt in agent mode that you find yourself running over and over again, or if you find yourself digging through past sessions to copy-paste your old successful prompts, it’s time to create a playbook. Playbooks are reusable plans that you can create and save for your most common and complex tasks. Instead of retyping detailed instructions every time, you can simply trigger a Playbook to get the job done. This is perfect for tasks like:

  • Generating weekly reports
  • Analyzing sales data
  • Creating marketing campaign assets
  • Onboarding new clients

Best practices and example playbooks can be found here.

To get started, select Playbooks from the left menu

From the Playbooks page select New playbook.

Define Your Playbook:

  • Name: Give your playbook a clear, descriptive name you'll easily recognize, like "Weekly Report Summary" or "Blog Post Social Media Pack."
  • Instructions: This is where you write the prompt you want to reuse. Be as detailed as you would be in a normal WRITER Agent chat!

đź’ˇ Pro-Tip: To make your playbooks even more powerful, you can include variables. For example, instead of naming a specific document, you can use a placeholder that you fill in each time you run the Playbook. For full instructions and best practices on how to get the most from playbooks please click here.

Finally select Create and your playbook will be made available in the Playbooks tab.

Using playbooks

Once your playbook is created you’ll see it in the list within the Playbooks tab. Here you can run, edit, and delete the playbook by selecting the three dot ••• icon next to a playbook. You'll also be able to create a routine from this playbook (more on this in a moment).

To launch a playbook from WRITER Agent, select + menu, then select the playbook you'd like from the menu. Once you select your playbook, WRITER Agent will fill in your prompt input field with the instructions. If your playbook contains variables you will be asked to complete these fields before you're able to run the agent.

Scheduling routines

Routines are automated triggers that run your playbooks on a recurring basis. You can set up a routine to run a playbook at a specific time, such as daily, weekly, or monthly. Each time a routine runs, it will create a new session in WRITER Agent.

From the Playbooks tab you can schedule a routine using the three dot ••• icon next to a playbook.

Additionally, you can select Manage WRITER Agent in the top right and then select the Routines tab.

You'll then be asked to name your routine then select how often it should run and when. Once you’ve set your time, select your timezone and then select Next.

If you've created a playbook you'll be able to select it from the list and save your changes. If you haven't created a playbook yet you will need to do that first. Please take note if you'd like your agent to run autonomously your playbook should not contain variables as these will need to be completed each time your playbook runs.

FAQs

Where can I access WRITER Agent?

WRITER Agent is available on the web app.

What is a session?

A session is an instance of WRITER Agent where the chat history and uploaded files will persist even if you leave the chat. If there is a set of files you’ll work with consistently, you can upload them into a session and return to it repeatedly.

Does WRITER Agent consider the entire chat history when responding to a request?

WRITER Agent will consider the full chat history for all options you select except when working with Knowledge Graph. If you select a Knowledge Graph within WRITER Agent, it will consider the most recent three chat back-and-forths between you and WRITER as context for what it generates next. If you do not want WRITER Agent to consider previous chat history, you can create a new session.

Are sessions shareable with teammates?

Yes you can share your session replay using the share icon in the top right. This will show another user your session but will not allow them to chat with and build upon the session you share.

How do you access previous sessions?

You can access previous sessions by going to the Sessions menu on the left as well as from the My Work section.

Can you delete previous sessions?

Yes, you can delete previous sessions.

What types of files can WRITER Agent handle?

  • Can parse text from these file types: xlsx/xls, csv, pptx/ppt, pdf, html, txt, srt, docx/doc, url, pasted text, past sources.
  • Can analyze .jpg and .png images.
  • Can parse text from tables but may not understand table formatting and layout.
  • Can parse PDFs but quality depends on how files were created. For example, if PDFs may have text that is an image, a file’s permission may not allow text to be copied, or the layout of a PDF’s layout may be hard for us to interpret (e.g., WRITER may struggle to render that a newsletter with multiple columns).

Can WRITER Agent parse images?

Yes WRITER Agent has image parsing capabilities.

Can WRITER Agent parse numerical data?

WRITER Agent can fetch/receive numbers from uploaded sources, but it cannot perform operations at this time. For example, if you uploaded a document that said, "Brickly has 100 employees, 43 of whom are women" you'd be able to ask WRITER Agent, "How many employees does Brickly have?" and "How many women are employed at Brickly?" but you couldn't ask "What percent of Brickly employees are women?" at this time.

Can WRITER Agent translate content?

WRITER Agent can translate content <500 words. For longer translation, use a translation prebuilt agent.

Are the files you upload available to teammates?

No, they are private to just you unless you share the session.

Does WRITER Agent store the files I upload?

We store the files you upload to WRITER Agent to our Google Cloud Platform (GCP).

Can WRITER Agent analyze text across multiple files?

Yes, WRITER Agent will be looking across all source files you upload by default.

Can you reference files you previously uploaded in a new session?

No.

I've uploaded a PDF but I'm getting an error message after I submit my prompt, "I apologize, but as an AI text-based model, I am unable to directly access or analyze attached files."

Double-check whether your PDF is encrypted. For security reasons, we cannot override the settings of your files. To check whether your PDF is encrypted:

  • Adobe Acrobat:
    • Windows: Choose the hamburger menu > Document Properties > Security - Are there restrictions under Document Restrictions Summary, e.g. Page Extraction?
    • MacOS: File > Document Properties > Security - Are there restrictions under Document Restrictions Summary, e.g. Page Extraction?
  • MacOS Preview:
    • Open the file. Click the (i) Info button in the toolbar. Select the lock to view security settings.

Where can I see the results of a scheduled Routine?


Each time a Routine runs, it creates a brand new session in WRITER Agent. You will find the completed task in your list of recent WRITER Agent sessions, titled with the name of the routine and the date it ran.

Can I edit a Playbook or a Routine after I've created it?

Yes! Simply navigate to the Playbooks or Routines tab, and you will see options to edit or delete any of the items you've created.

What happens if I delete a Playbook that is being used by a Routine?

If you delete a Playbook, associated routines will also be deleted.

Security, transparency & other guardrails

How do we mitigate security, safety, and privacy risks given the level of autonomy and access of WRITER Agent?

WRITER Agent is designed with a multi-layered approach to mitigate risks associated with its autonomy, and backed by the security, reliability, and control provided by the WRITER platform.

Security Controls:

  • Isolated Sandbox Environment: Operates within a secure, containerized sandbox that is completely isolated from the host infrastructure and other user sessions.
  • Temporary Workspace: All files, data, and running processes are temporary and are permanently destroyed at the end of each session. This design prevents the persistence of malware or sensitive data.
  • Restricted Network Access: Outbound network connections are monitored and filtered, limiting it to standard web protocols (HTTP/S) and preventing malicious activities like network scanning or participation in DDoS attacks.

Safety Protocols:

  • Behavioral Guardrails: The underlying language model has built-in safety filters to prevent the generation of harmful, illegal, or unethical content.
  • Human-in-the-Loop Design: For critical actions, such as deploying applications or deleting files, the agent is programmed to stop and request explicit confirmation from the user. The user retains ultimate control over significant operations.
  • Resource Limiting: All processes are strictly limited in their use of CPU, memory, and execution time to prevent system instability or resource abuse.

Privacy Protections:

  • Stateless Sessions: WRITER Agent has no long-term memory and does not retain any data between sessions. Each interaction begins from a clean slate.
  • User-Controlled Data: WRITER Agent only has access to data that the user explicitly provides during an active session. It has no access to a user's local machine or files.
  • Encrypted Communication: All data transmitted between the user and the agent is protected with industry-standard HTTPS encryption.
Did this answer your question? Thanks for the feedback There was a problem submitting your feedback. Please try again later.