autonomous-learning

name	autonomous-learning
description	Dialectical reasoning framework for self-improvement through pattern detection
version	1.0.0
triggers	["error patterns detected","repeated corrections","high novelty scores"]

Autonomous Learning

Learn from error patterns and generate improvements to prevent recurrence.

Pattern Detection

1. Repeated correction: User corrects the same type of mistake twice
2. Tool failure: Same tool fails with same error pattern
3. Approach rejection: User rejects the same approach category twice

Learning Loop

1. Detect: Notice the pattern (correction, failure, rejection)
2. Analyse: What went wrong and why?
3. Generalise: Is this specific to this context or universal?
4. Record: Save the insight to memory for future sessions
5. Apply: Use the insight in the current session

Output

When a learning insight is generated:

[LEARNING] Pattern: [description]
Insight: [what to do differently]
Scope: [project-specific | universal]