Structured Data
Capability Metadata
| Field | Value |
|---|---|
| ID | CAP-102 |
| Version | v1.27 |
| Domain | Specialist Domains |
| Group | Knowledge Packs |
| Status | Documented |
| Related | CAP-101, CAP-103, CAP-104 |
| Scaffold levels | full-scaffold, partial, minimal, none |
| Hint delivery | on-request, automatic, time-triggered |
Configuration Reference
Scaffolding levels:
full-scaffold — step-by-step guidance with worked examples (novice)
partial — hints on first error; answer revealed after N attempts (practitioner)
minimal — reference links only; no hints (expert)
none — bare exercise, no support; for mastery verification
Hint delivery modes: on-request (user asks), automatic (on first error), time-triggered (after N seconds)
Exercise: Write an AGC routine to load constant 0o42 into A
Full scaffold (novice):
Step 1: Use CAF instruction to load from Fixed memory
Step 2: CAF CONSTANT loads the value at address CONSTANT
Step 3: Define CONSTANT = 0o42 with EQUALS directive
Code: CONSTANT EQUALS 0o42
CAF CONSTANT
Partial scaffold (practitioner):
Hint: CAF loads from fixed memory. EQUALS defines constants.
[Show answer after 3 attempts]
Minimal scaffold (expert):
No hints. Reference: AGC Instruction Set manual §3.2Technical Specifications
Scaffold selection: driven by mastery level from Knowledge Adaptation (CAP-101)
Worked example structure: problem statement → steps → solution → explanation → variation
Hint removal schedule: automatic reduction as concept_score increases — full-scaffold → partial after score > 0.6, partial → minimal after score > 0.8
Competence evaluation: exercise results feed back to knowledge patterns (CAP-100)
Overview
Scaffolding makes the difference between a learner successfully completing an exercise and abandoning it in frustration — providing just enough support without preventing the productive struggle that builds genuine competence. Knowledge Scaffolding wraps knowledge pack exercises with the right level of support, matched to the learner's current mastery level via Knowledge Adaptation (CAP-101).
Four scaffold levels cover the full range from complete beginner to verified expert. Full scaffolding provides step-by-step guidance and worked examples — a novice learning AGC assembly needs to see exactly how to define a constant and load it into the accumulator before attempting it independently. Partial scaffolding provides hints on first error and reveals the answer after multiple failed attempts. Minimal scaffolding provides only reference links. No scaffolding is for mastery verification — the learner demonstrates they can complete the exercise without any support.
Scaffolding is progressive: as a learner's concept scores improve (tracked by Knowledge Patterns, CAP-100), the scaffold level automatically reduces. A learner starts with full scaffolding on a new exercise. After demonstrating consistent success, the system silently removes the worked example and provides only hints. After further success, hints disappear. The learner never explicitly requests this reduction — it happens automatically as competence is demonstrated.
History
Full scaffolding was added after observing high abandonment rates on complex exercises — learners with no context for AGC arithmetic needed step-by-step guidance before they could attempt independent work. The original knowledge pack exercises were bare problem statements: "Write an AGC routine to load 0o42 into the accumulator." Without any scaffolding, learners unfamiliar with AGC instruction syntax had no way to start.
The progressive reduction mechanism was inspired by the educational concept of "fading" — gradually removing support as competence develops. The automatic reduction (triggered by concept_score thresholds rather than explicit learner requests) was specifically designed to prevent learners from becoming scaffolding-dependent. If learners had to explicitly ask for reduced scaffolding, few would — the default would be to keep the support available "just in case." Automatic fading makes the reduction inevitable as competence grows, mimicking how a good teacher naturally gives students more independence as they demonstrate they're ready.
Key Concepts
Scaffolding in construction is temporary support that holds a building up while it's being built — it's essential during construction but gets removed once the structure can stand on its own. Educational scaffolding works the same way: temporary support that enables learning but gradually disappears as the learner becomes independent.
Knowledge Scaffolding provides this temporary support automatically. You don't have to tell it when you're ready for less help — it watches your exercise performance and reduces the support as you demonstrate you don't need it anymore. The goal is for you to eventually complete exercises without any scaffolding at all, which is when you're genuinely ready for Knowledge Promotion (CAP-103) to turn your competence into a permanent skill.
Usage Examples
Example 1: Defining scaffolded exercises in a knowledge pack
# In knowledge pack exercises section
exercises:
- id: load-constant-into-accumulator
title: "Load a constant value into the AGC accumulator"
scaffolding:
full:
steps:
- "Use CAF (Clear and Add Fixed) instruction"
- "Define CONSTANT with EQUALS directive"
- "CAF CONSTANT loads the value"
solution: |
CONSTANT EQUALS 0o42
CAF CONSTANT
partial:
hint: "CAF loads from fixed memory. EQUALS defines constants."
show_answer_after_attempts: 3
minimal:
reference: "AGC Instruction Set manual §3.2"Example 2: Scaffold level automatically reduces
Session 1: concept_score['CAF'] = 0.4 → full-scaffold shown
Session 3: concept_score['CAF'] = 0.65 → partial scaffold shown (auto-reduced)
Session 6: concept_score['CAF'] = 0.82 → minimal scaffold shown (auto-reduced)
Session 8: concept_score['CAF'] = 0.91 → no scaffold (mastery verification mode)