Repository is archived
No release in over 3 years
Collins' Cognitive Apprenticeship model for brain-modeled agentic AI: guided learning through modeling, coaching, scaffolding, articulation, reflection, and exploration
2005
2006
2007
2008
2009
2010
2011
2012
2013
2014
2015
2016
2017
2018
2019
2020
2021
2022
2023
2024
2025
2026
 Dependencies

Development

 Project Readme

lex-cognitive-apprenticeship

Collins' cognitive apprenticeship model for brain-modeled agentic AI — six pedagogical methods track skill acquisition toward mastery.

What It Does

Models how agents learn complex skills through apprenticeship relationships. Based on Collins' cognitive apprenticeship theory, each learning session uses one of six methods (modeling, coaching, scaffolding, articulation, reflection, exploration), each with different learning multipliers. Mastery increases per session until it crosses the graduation threshold (0.85).

Core Concept: Six Methods and Their Multipliers

modeling     -> base gain (0.08)     # observe expert performing
coaching     -> 1.5x gain (0.12)     # guided practice with feedback
scaffolding  -> base gain (0.08)     # supported structure
articulation -> base gain (0.08)     # explaining what you know
reflection   -> base gain (0.08)     # comparing to expert model
exploration  -> 2.0x gain (0.16)     # unsupported problem solving (biggest jump)

Usage

client = Legion::Extensions::CognitiveApprenticeship::Client.new

# Register a new apprenticeship
ap = client.create_apprenticeship(
  apprentice: :lex_agent,
  mentor: :senior_engineer,
  domain: :infrastructure,
  skill: :incident_response
)

# Run a coaching session
client.conduct_apprenticeship_session(
  apprenticeship_id: ap[:apprenticeship][:id],
  method: :coaching
)
# => { mastery_delta: 0.12, mastery: 0.37, phase: :novice }

# Get a recommendation on what to do next
client.recommend_apprenticeship_method(apprenticeship_id: ap[:apprenticeship][:id])
# => { recommended_method: :scaffolding, reason: "low mastery — structure needed" }

# Run an independent exploration session at higher mastery
client.conduct_apprenticeship_session(
  apprenticeship_id: ap[:apprenticeship][:id],
  method: :exploration
)
# => { mastery_delta: 0.16, mastery: 0.53, phase: :apprentice }

# Check for graduates
client.graduated_apprenticeships
# => apprenticeships where mastery >= 0.85

Integration

Pairs with lex-agency (a graduated apprenticeship maps directly to record_mastery — the strongest Bandura efficacy source). Pairs with lex-memory (session outcomes stored as procedural traces) and lex-prediction (predict time to graduation from mastery trajectory).

Development

bundle install
bundle exec rspec
bundle exec rubocop

License

MIT