CurricuLLM LogoCurricuLLM
In the ClassroomFeaturesPricingTraining HubDevelopersFAQ
3.9 Using the Progression Agent
Training Hub3. Teacher Guides3.9 Using the Progression Agent

3.9 Using the Progression Agent

How to use @ tagging and the + button to incorporate student progression data into your prompts for personalised differentiation, reporting, and intervention planning.

The progression agent is a powerful feature that lets you bring real student data directly into your CurricuLLM conversations. By using @ tags or the + button to attach students or classes to your prompt, you get access to all their progression data—making your teaching truly responsive to where students actually are, not where you think they might be.

This section shows you practical workflows for using progression data to differentiate work, write accurate reports, plan interventions, and group students effectively.


What is the Progression Agent?

The progression agent is CurricuLLM's way of making every conversation grounded in real student learning data.

How it works

When you're writing a prompt, you can:

  • Type @ followed by a student name to tag an individual student
  • Type @ followed by a class name to tag an entire class
  • Click the + button in the chat interface and select students or classes from a list

Once attached, CurricuLLM has access to:

  • Learning progression levels across curriculum areas
  • Recent assessment results and patterns
  • Strengths and areas needing development
  • Previous work history and growth trends
  • Identified learning gaps or misconceptions

When to use it

Use the progression agent whenever you need to:

  • Differentiate activities based on actual student levels, not assumptions
  • Write report comments that accurately reflect individual progress
  • Create groups for guided teaching based on real data
  • Plan interventions for students struggling in specific areas
  • Design extensions that genuinely challenge advanced learners

The progression agent makes personalisation accurate, fast, and grounded in evidence.


Workflow 1: Differentiating Activities Based on Student Progression

The scenario: You're planning tomorrow's writing lesson on persuasive techniques. Your Year 8 class has a wide range of writing abilities. Instead of guessing who needs what level of work, you want to use actual progression data to create three appropriately pitched versions.

What you do:

  • Open CurricuLLM and start a new chat.
  • Type @Year8English or click the + button and select your Year 8 English class.
  • Say: "I'm teaching persuasive writing tomorrow. Based on my class's progression data, create three versions of a persuasive writing task:
    • Version A: For students working below expected level, with scaffolding and sentence starters
    • Version B: For students at expected level, with guided structure
    • Version C: For students above expected level, with complex argumentation challenges

Include which students should receive each version based on their current writing progression."

  • CurricuLLM analyses the progression data and generates three versions.
  • You also get a suggested student grouping showing exactly who should get each version.
  • Review, adjust if needed, and print or share digitally.

What you get:

  • Three versions of the task, each pitched accurately to real student levels
  • Clear rationale for why each student gets their version
  • Scaffolding that addresses specific gaps visible in the data
  • Extension work that targets actual next-level skills
  • Suggested groupings you can use immediately

Time saved: Creating differentiated versions from scratch with no data: 60+ minutes of guessing. Using progression agent: 5 minutes with accuracy.

How to use it in class:

  • Students receive their version without fanfare—it's just "today's task"
  • Each version challenges students appropriately
  • You can check in with specific students knowing exactly what they're working on
  • Monitor who succeeds to update your understanding of their level
  • Everyone works toward the same learning goals, just with different pathways

Teacher tips:

  • The progression agent shows you which students are borderline between versions—use your classroom judgment for those cases
  • You can ask follow-up questions like "Why did you place Marcus in Version B?" to understand the reasoning
  • If you disagree with a placement, you can override it—the data informs your decisions, it doesn't make them
  • Save successful differentiated sets to reuse the structure with different topics

Workflow 2: Writing Data-Informed Report Comments

The scenario: It's report writing time. You have 28 students to write comments for in English. Instead of relying purely on memory and recent impressions, you want comments that accurately reflect each student's progression across the term, with specific reference to their growth areas and next steps.

What you do:

  • Open CurricuLLM and start a chat for report writing.
  • For the first student, type @EmmaJohnson (or use the + button to select her).
  • Say: "Write a 100-word report comment for Emma in Year 7 English. Reference her progression data, highlight specific strengths, note areas for development, and suggest clear next steps. Make it parent-friendly and encouraging but honest."
  • CurricuLLM generates a comment grounded in Emma's actual data.
  • Read it, adjust the tone or wording to match your voice, and copy to your report system.
  • Continue with the next student.

What you get:

  • Comments that reference specific progression markers and growth
  • Accurate identification of strengths grounded in data, not impressions
  • Development areas that match actual gaps in learning progression
  • Next steps that logically follow from current achievement
  • Consistent structure but genuinely personalised content

Time saved: Writing 28 individualized comments from scratch: 4-5 hours. With progression agent: 45-60 minutes including review and adjustment.

How to use the results:

  • Copy comments into your school's reporting system
  • Adjust wording to match your voice and school's expectations
  • Add personal anecdotes or classroom observations to complement data
  • Use the comments as conversation starters for parent-teacher meetings
  • Parents receive accurate, evidence-based feedback

Teacher tips:

  • Be specific in your prompt about report comment style (encouraging, developmental, formal, conversational)
  • Mention if your school has specific requirements (character limits, structure, terminology)
  • The progression agent prevents the "I can't remember what they did in Week 3" problem
  • You can generate comments in batches (e.g., 5 students at a time) if you prefer more control
  • Always add at least one personal touch that only you as the teacher would know

Key Principles of Using the Progression Agent

1. Data makes differentiation accurate, not guesswork

Without data, differentiation is based on impressions, recent work, or gut feeling. With progression data, you know exactly where each student is and what they need next. This transforms differentiation from hopeful to precise.

2. Speed enables responsive teaching

When you can analyse 28 students' progression data and create appropriate groups or resources in 5 minutes, you can respond to your class as they are today, not as they were last term. Teaching becomes more agile and responsive.

3. Privacy and data protection built-in

All progression data is protected under strict privacy regulations. Students and parents can see what data is held. Only authorized teachers can access student progression information. Data is never used for purposes beyond supporting learning.

4. Teacher judgment remains central

The progression agent provides evidence and suggestions, but you make all final decisions. You know your students in ways data never can—their confidence, motivation, social dynamics, and daily variability. Use progression data to inform your professional judgment, not replace it.

5. Accuracy improves teaching efficiency

When differentiation is accurate, fewer students are frustrated (work too hard) or bored (work too easy). This means less behaviour management, better engagement, and more actual learning happening. Accurate targeting saves time in the long run.


Getting Started with the Progression Agent

Start with one student:

Before using progression data for your whole class, practice with one student. Tag them, ask a question about what they need, and see how the progression agent responds. Build your confidence with the feature.

Try a simple use case first:

Start with something straightforward like creating two versions of a worksheet (challenge and support) for one lesson. This lets you see how progression data informs differentiation without overwhelming complexity.

Build to small groups, then whole class:

Once comfortable, try creating groups of 3-5 students for guided teaching. When that feels natural, move to whole-class reporting or differentiation where you're using data for all students at once.

Save effective prompts:

When you write a prompt that gives you exactly what you need, save it. Next time you need similar support, use the same prompt structure and just change the topic or student names.

Troubleshooting common issues:

  • If progression data seems outdated: Check that recent assessments have been entered into the system.
  • If student names don't appear when you type @: Make sure you're using the correct class list or check with your admin that student data is loaded.
  • If suggestions don't match your knowledge: Trust your classroom experience and adjust. The data shows patterns, but you know daily realities.
  • If groupings seem off: Remember data shows progression levels, not behaviour or social factors. Adjust groups to suit your classroom needs.

Everyday Analogy

Using the progression agent is like having detailed medical records when treating patients. Instead of asking "How do you feel?" and guessing symptoms, a doctor looks at test results, medical history, and vital signs before deciding treatment. This doesn't replace the doctor's expertise—it enhances it. The doctor still makes decisions, but with evidence.

The progression agent works the same way. Instead of relying only on memory and impressions, you have clear data about where each student is in their learning. You still make all the teaching decisions—what to teach, how to group, what support to give—but now you're making those decisions with evidence, not guesswork. The result is more accurate teaching that meets students exactly where they are.

Previous
3.8 Using Student Emulation
Next
3.10 Using Diagrams to Support Learning
CurricuLLM Logo
CurricuLLM

AI for schools

Product

FeaturesPricingDevelopersUse CasesFAQ

Company

About usPrivacy policyStatusContact

Resources

Terms of useSupportTraining hubBlog