Documentation Index
Fetch the complete documentation index at: https://docs.sxth.ai/llms.txt
Use this file to discover all available pages before exploring further.
Learning Adapter
The Learning adapter helps users develop skills with progress tracking and stuck detection.
Installation
from sxth_mind import Mind
from examples.learning import LearningAdapter
mind = Mind(adapter=LearningAdapter())
Identity Types
| Type | Description | Teaching Style |
|---|
conceptual | Needs to understand “why” | Explain theory first |
hands_on | Learns by doing | Jump to examples |
structured | Follows curriculum | Step-by-step progression |
explorer | Self-directed discovery | Provide resources, let them lead |
Journey Stages
exploring → foundations → practicing → applying → deepening
↓ ↓
stuck ←────────────────────────┘
| Stage | Tone | Focus |
|---|
exploring | Welcoming | Understand goals, assess level |
foundations | Patient | Core concepts, fundamentals |
practicing | Encouraging | Exercises, repetition |
applying | Challenging | Projects, real-world use |
deepening | Collaborative | Advanced topics, edge cases |
stuck | Supportive | Identify blocker, simplify |
Pattern Detection
The Learning adapter tracks:
- Exercises completed — Practice volume
- Projects completed — Applied learning
- Stuck indicators — Repeated questions, frustration signals
- Learning velocity — Progress rate
- Preferred format — Examples vs. theory
Nudge Templates
| Nudge | Priority | Trigger |
|---|
practice_reminder | 5 | No practice in 3+ days |
stuck_support | 8 | Detected frustration or repeated asks |
milestone | 4 | Completed exercise set or project |
next_level | 5 | Ready to advance |
Example Usage
import asyncio
from sxth_mind import Mind
from examples.learning import LearningAdapter
async def main():
mind = Mind(adapter=LearningAdapter())
# Learning Python
await mind.chat("user_1", "I want to learn Python", project_id="learn_python")
await mind.chat("user_1", "What are list comprehensions?", project_id="learn_python")
await mind.chat("user_1", "I don't understand, can you show an example?", project_id="learn_python")
# The adapter notices preference for examples
state = await mind.get_state("user_1", project_id="learn_python")
print(f"Identity type: {state['user_mind'].get('identity_type')}")
# → Likely "hands_on" based on asking for examples
asyncio.run(main())
Multi-Topic Support
Track multiple learning topics:
await mind.chat("user_1", "Learning Python basics", project_id="learn_python")
await mind.chat("user_1", "Started a React course", project_id="learn_react")
await mind.chat("user_1", "Practicing SQL queries", project_id="learn_sql")
# Each topic has its own progress and stage
Stuck Detection
The adapter detects when users are stuck:
# These signals trigger "stuck" stage:
# - Repeated similar questions
# - Phrases like "I don't understand", "confused", "stuck"
# - Long gaps in progress
# - Decreasing momentum
# Response adapts:
# - Simplify explanations
# - Offer alternative approaches
# - Break down into smaller steps
Context Data
The adapter stores in project_mind.context_data:
{
"topic": "python",
"current_module": "loops",
"exercises_completed": 12,
"projects_completed": 1,
"stuck_count": 2,
"preferred_format": "examples",
}
