Story copilot is a conversational AI interface in Tines to build, understand, and optimize stories. This article outlines key recommendations to help you use Story copilot.
Read our Introduction to Story copilot article here.
There are two main modes in Story copilot:
Ask - Ask anything about the story on screen or general Tines queries
Build - Build, edit or debug your workflow
Ask mode
Ask mode is designed purely to help you understand and assess existing Tines stories or ask general Tines queries. Its primary goal is to provide clarity and insight into any story on your screen.
Example: Need to understand a complex story
Prompt: “This is a big story, can you summarize it for me?”
Result: Story copilot will summarize the story, give you information such as its purpose, overview, key components, integration points, any issues and some notable features being used.
Example: Reviewing a critical story for maintenance.
Prompt: “Wanting a health check on my story. Are there any updates I should make?”
Result: Story copilot will analyze the configuration of the actions in the story, check credentials, check action logs for errors, and suggest specific changes.
Tips
Query safely: Use Ask mode when querying a story to prevent accidental edits.
Access: if you don't have editor access to storyboards, you will only see Ask mode.
Build mode
Build mode is the comprehensive interface for all workflow construction, modification, and troubleshooting. It functions as the umbrella for three core capabilities: build, edit, and debug.
Note: Build, edit and debug are not separate modes you have to select, they all live under Build mode.
Build
This allows you to create a story from scratch.
Example: Create a story that triages an alert and sends a notification if its critical
Prompt: “I have an alert that follows this format:
{
“alert”’ {
“user_name”: “tino”,
“ip”: “194.88.246.242”,
“email”: “tino@tinesdemos.com:,
“device”: “MacBook Pro 16”
}}
Build out a story that receives the alert, triages it, and emails the team if it’s critical”
Result: Story copilot will analyze the request to plan the workflow. It may ask you some confirming questions first like “Which threat intelligence service would you like to use for IP reputation checks?”. It will search existing templates to see if one exists already. If there is one, it will give you a diagram of the workflow just before it starts to build it. Once it has built the story, it will also test the workflow to make sure it works. Story copilot will then summarize the story for you and you will have the option to review and accept or refuse the suggested story.
Example: A beginner user attempting their first workflow.
Prompt: “I’m a new builder and need help getting started. I want to build a flow that checks a list of IPs against a threat intelligence feed and blocks them if they are malicious.”
Result: Story copilot ask a few confirming questions before drafting the workflow, explaining the steps for a new user, and possibly offering to guide them through the configuration.
Edit
This allows you to make changes to an existing story.
Example: Adding a final notification step to a story.
Prompt: “When the entire vulnerability remediation story finishes running successfully, can you add a step to post a summary of the results to the #security-ops Slack channel with the tag @security-lead?”
Result: Story copilot will add a new Slack Send Message Action at the end of the workflow, configure it to post the summary, and set up the necessary conditional logic.
Example: Preparing a finished story for team handover or production.
Prompt: “I finished building the offboarding story. Can you rename and write descriptions for all the actions to improve clarity?”
Result: Story copilot will rename all actions with clear, descriptive titles (e.g., 'HTTP Request to HR System' to 'Fetch Employee Manager') and add a concise description to each action explaining its purpose.
Debug
This can help you figure out why a story isn’t working.
Example: Messages not being sent to Slack
Prompt: “Why isn’t my slack message sending?”
Result: Story copilot will debug the issue for you and when it finds the issue, it will communicate it back to you and explain why it is not working. Next, it will fix the issue and give you the option to review and accept or refuse the change.
Example: A story is failing with an unknown error.
Prompt: “My story isn't running the way I want it to. I expected it to block the IP address, but it only sent the email. Can you help me figure out what’s wrong?”
Result: Story copilot will analyze the story's execution history, identify the point of failure or divergence, and explain the issue, offering a fix.
Tips
Be specific and concise: Provide succinct but detailed prompts to ensure Story copilot accurately understands your goal.
Review all changes: Always review and approve Story copilot's suggested actions to maintain full control over your workflow.
Read this article on Writing effective prompts for more help.
Using your own style guides
To build or customize stories using your team's internal style guides and best practices, you can create a Story copilot guidelines resource.
This article will help you create resources in Tines.
To leverage your resource, simply ask Story copilot to reference the resource when you are building. This makes team-wide use and updates easy as your internal best practices evolve.
Read more about Story copilot here.