Contesimal: Agent Library - Understand

Description

The Agent Library - Understand feature lets you use Rumi (Contesimal's AI agent) to analyze your audience's values, perspectives, sophistication levels, intent, and learning orientation based on your content library. You can choose from four analysis types: Audience Values & Perspectives, Audience Sophistication Assessment, Audience Intent & Learning Orientation, and Audience Overview Insights.

Each analysis type provides specific tasks tailored to understanding different aspects of your audience (e.g., "Perspective Consistency Report", "Cognitive Load Profile Report", "Intent Consistency Report", "Create Audience Purpose Assessment"). After selecting your analysis type and task, providing optional context, and selecting a project, Rumi processes your request and generates audience insights and reports that you can refine and use in your Creation Canvas.

Step-by-Step: 

The step-by-step instructions, screen descriptions, and UI details in this section are examples only. They are intended to help you understand how the feature works and how to complete the flow. Screens, labels, and layout may change in future product updates.

1. Access Agent Library - Understand 

• Navigate to the Agent Library section In left navigation menu
• Click Understand section

2.  Select Understand Type

• On the "Select an option" screen, choose one of the four analysis types:
  • Audience Values & Perspectives
    • Analyze audience values, worldviews, and perspective alignment
  • Audience Sophistication Assessment
    • Assess audience cognitive level, knowledge familiarity, and mental model assumptions
  • Audience Intent & Learning Orientation
    • Analyze audience intent, learning preferences, and consumption patterns
  • Audience Overview Insights
    • General audience insights and assessments
• Click on your desired analysis type
• You'll be taken to the task selection screen

3.  Select Task

• After selecting an analysis type, you'll see a "Select a task" screen with specific task options for that analysis type
• For Audience Values & Perspectives, available tasks include:
  • Perspective Consistency Report
    • Evaluates how aligned or fragmented audience value signals are across the analyzed document set
  • Language Density Report
    • Quantifies prescriptive or belief-driven language to infer how strongly values are asserted
  • Perspective Alignment Report
    • Measures whether the content reflects a conventional, balanced, or contrarian worldview relative to common industry norms
  • Trade-Off Preference Report
    • Surfaces implicit prioritizations (such as speed vs. quality or scale vs. control) expressed through comparative language
  • Value Signal Frequency Report
    • Identifies recurring value indicators (e.g., efficiency, quality, autonomy, growth) embedded in content summaries
• For Audience Sophistication Assessment, available tasks include:
  • Sophistication Stability Report
    • Assesses whether the audience cognitive level remains stable or varies significantly across content selections
  • Concept Reuse & Familiarity Report
    • Analyzes repeated concepts and phrasing to infer what knowledge is treated as familiar vs. newly introduced
  • Cognitive Load Profile Report
    • Estimates the mental effort required to consume the content based on idea density, abstraction shifts, and inferential gaps
  • Mental Model Assumption Report
    • Identifies whether the content assumes prior frameworks, heuristics, or first-principles thinking from the audience
  • Concept Depth Distribution Report
    • Measures the complexity and layering of ideas to infer whether the audience skews beginner, intermediate, or advanced
• For Audience Intent & Learning Orientation, available tasks include:
  • Intent Consistency Report
    • Evaluates how consistent audience intent signals are across the library or within
  • Exploration vs. Resolution Orientation Report
    • Measures whether content invites open-ended exploration or pushes toward concrete answers and conclusions
  • Outcome Expectation Report
    • Surfaces the type of value the audience expects (knowledge gain, application, mindset shift, or validation) as inferred across your library
  • Learning Mode Preference Report
    • Determines whether content is primarily conceptual, procedural, narrative, or principle-driven based on language and structure patterns
  • Primary Consumption Intent Report
    • Identifies the dominant intent (e.g., learning, exploration, inspiration, entertainment) inferred from semantic similarity across content summaries
• For Audience Overview Insights, available tasks include:
  • Create Audience Purpose Assessment
    • Break down the main purposes underlying your core audiences reason for listening or watching
  • Get Strategy for Measuring Performance
    • Report on how to measure your content's performance at a high-level
  • Define Audience Problems to Solve
    • Address common questions, challenges, or aspirations your audience faces
  • Generate an Audience Category Report
  • Discover key categorical trends for your main audience
• Click on the task card that matches your needs

4.  Add Additional Context (Optional)

On the "Would you like to add additional context?" screen, you have three options:

• "Add from your Library" - Select documents from your existing content library
  • Select one document from the Content Library and click "Add Content" button OR
  • Search for a document from the Content Library and click "Add Content" button 
  • Click the X icon on selected document to remove it and add another one
• "Add a URL" - Provide a web URL for Rumi Agent to reference
• "Upload a document" - Upload a new document from your computer
  • Suppported formats: PDF, DOC, DOCX, TXT, 
  • Max size 480 KB
• Click the X icon  to remove a context item and upload a new one
• Click "Next" to proceed, or "Skip" if you don't want to add context

5.  Select or Create Project

On the "Select or create project" screen, choose where to save your research:

• Select an existing project from the dropdown menu
• Use the search bar to find a specific project
• Click "+ Create new Project" to create a new project
• Click "Next" button

6.  Rumi Agent Processes Your Request

• You'll see: "Rumi is working on your request" screen
• A progress bar will show at the bottom
• Wait for the Agent to complete the research (typically takes a few seconds to a few minutes)

7.  Review Results on Creation Canvas

• Once Rumi Agent completes the research, you'll be automatically redirected to the Creation Canvas
• Your research results will be displayed in a document format
• You can review, edit, refine, delete or share your creation

Troubleshooting Guide

A. Analysis Type Options Don't Appear

• Issue: After accessing Understand, the analysis type selection screen is blank or options don't load.

• Fix: 

  • Refresh the page and wait a few seconds for options to load

  • Check your internet connection - ensure you have a stable connection

  • Clear browser cache and try again

  • Try a different browser (Chrome, Edge, Firefox) if the issue persists

  • If options consistently don't appear, contact Support and mention that marketing content creation types aren't loading

B. Task Cards Don't Appear

• Issue: After selecting an analysis type, the task selection screen is blank or task cards don't load.

• Fix:

  • Refresh the page and wait a few seconds for channels to load

  • Check your internet connection - ensure you have a stable connection

  • Clear browser cache and try again

  • Try selecting the analysis type again

  • If tasks consistently don't appear, contact Support and mention which analysis type you selected

C. Task Cards Don't Respond to Clicks

• Issue: Clicking on a task card (e.g., "Perspective Consistency Report") doesn't proceed to the next step.

• Fix: 

  • Refresh the page and try clicking the task card again

  • Wait a moment - there may be a brief delay before navigation

  • Try clicking directly on the card text rather than the edges

  • Check browser console for any JavaScript errors

  • Try a different browser if the issue persists

  • If clicking consistently doesn't work, contact Support with the specific content creation type and task you're trying to select

D. Can't Find Documents in Library Search

• Issue: When adding context from Library, you can't find the documents you're looking for.

• Fix:

  • Use the search bar - type keywords from the document title

  • Check document names - ensure you're searching for the correct title

  • Scroll through the list - there may be more documents below the visible area

  • Use filters - click the filter icon to narrow down document types

  • Check if documents are processed - ensure your documents have finished processing in the Library

  • If you still can't find documents, try uploading them directly using "Upload a document"

E. Selected Context Items Don't Appear or Can't Be Removed

• Issue: After selecting documents from Library, they don't show as tags, or you can't remove them with the X icon.

• Fix:

  • Refresh the page and try selecting documents again

  • Click the X icon directly on the tag to remove it

  • Check if documents were actually selected - look for checkmarks in the modal

  • Try selecting fewer documents - there may be a limit on context items

  • If context items consistently don't appear, contact Support and mention the context selection issue

F. Project Dropdown Doesn't Open or Shows No Projects

• Issue: The "Select or create project" dropdown doesn't open, or no projects are listed.

• Fix:

  • Click the dropdown arrow to ensure it opens

  • Use the search bar within the dropdown to find projects

  • Create a new project by clicking "+ Create new Project"

  • Refresh the page and try again

  • Check if you have existing projects - you may need to create one first

  • If the dropdown consistently doesn't work, contact Support and mention the project selection issue

G.  "Add from your Library" Modal Doesn't Open

• Issue: Clicking "Add from your Library" doesn't open the document selection modal.

• Fix:

  • Refresh the page and try clicking the button again

  • Check browser pop-up blockers - ensure they're not blocking the modal

  • Try a different browser (Chrome, Edge, Firefox) or use an incognito/private window

  • Check browser permissions - ensure JavaScript is enabled

  • If the modal consistently won't open, try the other context options (Add URL or Upload document) instead

H. Agent Takes Too Long or Gets Stuck

• Issue: The "Rumi is working on your request" message persists for an unusually long time, or the progress bar doesn't advance.

• Fix:

  • Wait a few minutes - complex research requests may take longer

  • Check your internet connection - ensure you have a stable connection

  • Refresh the page - sometimes a refresh can resolve temporary issues

  • Check if the research completed - try navigating to Creation Canvas manually to see if results are there

  • Try a simpler research request - very complex or broad requests may take longer

  • If Rumi remains stuck for more than 10-15 minutes, contact Support with details about your research request

I. Not Redirected to Creation Canvas After Research Completes

• Issue: Agent appears to have finished, but you're not automatically taken to the Creation Canvas.

• Fix:

  • Manually navigate to the Creation Canvas section

  • Check your project - look in the project you selected to see if research results are there

  • Clear browser cache and cookies, then try a new research request

  • Check browser settings - ensure no extensions are blocking redirects

  • Try a different browser if the issue persists

  • If automatic redirection consistently fails, contact Support and mention the redirect issue

J. Create Results Are Incomplete or Incorrect

• Issue: The create results on Creation Canvas don't match what you requested, or information seems wrong.

• Fix:

  • Review your instructions - ensure they were clear and specific

  • Check your context - verify that the documents/URLs you added are relevant

  • Try refining the research - use the AI Research panel to ask follow-up questions

  • Add more specific instructions - try running the research again with more detailed instructions

  • Check the research type and option - ensure you selected the correct research category

  • If results are consistently incorrect, contact Support with details about what you requested vs. what you received

K. Can't Add URL or Upload Document for Context

• Issue: The "Add a URL" or "Upload a document" options don't work when clicked.

• Fix:

  • Refresh the page and try clicking the option again

  • For URL: Ensure you're entering a valid, accessible web address

  • For Upload: Check file size limits and supported formats (PDF, DOCX, TXT, etc.)

  • Try a different browser if the issue persists

  • Check browser permissions - ensure file access is allowed

  • If these options consistently don't work, try using "Add from your Library" instead, or contact Support