Repository is archived
No release in over 3 years
Temporal planning horizon modeling for brain-modeled agentic AI — dynamic expansion/contraction based on stress, success, and construal level theory
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-horizon

Temporal planning horizon engine for brain-modeled agentic AI in the LegionIO ecosystem.

What It Does

Models how far ahead an agent can effectively plan. The horizon (0.0 to 1.0) represents planning depth and can be expanded through deliberate effort or contracted under stress. Stress narrows the effective horizon without touching the underlying capacity — stress relief fully restores effective reach. Projections are forward-looking cognitive estimates at a given temporal distance; only projections within the effective horizon are considered actionable. Confidence in projections decays with temporal distance. A background actor runs every 60 seconds to sample horizon status.

Usage

require 'legion/extensions/cognitive_horizon'

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

# Check current horizon state
client.get_horizon
# => { success: true, current: 0.5, effective: 0.5, stress: 0.0, label: :moderate }

# Expand planning depth
client.expand_horizon
# => { success: true, before: 0.5, after: 0.58, label: :moderate }

# Add a forward projection
client.add_projection(
  content: 'complete API auth layer',
  domain: :engineering,
  temporal_distance: 0.3,
  confidence: 0.8
)

# Apply stress (narrows effective horizon)
client.apply_stress(amount: 0.5)
# => { success: true, stress: 0.5, effective_horizon: 0.505 }

# See which projections are still within reach
client.projections_within_horizon
# => { success: true, projections: [...] }

# Relieve stress to restore full effective horizon
client.relieve_stress(amount: 0.5)

Development

bundle install
bundle exec rspec
bundle exec rubocop

License

MIT