lex-mental-time-travel

Autonoetic temporal self-projection for LegionIO agents. Part of the LegionIO cognitive architecture extension ecosystem (LEX).

lex-mental-time-travel enables an agent to project itself backward (retrospection) and forward (prospection) in time by constructing journeys with waypoints. Each waypoint's vividness decays with temporal distance. A 10% confabulation rate introduces realistic memory construction errors. The agent's autonoetic level — its sense of mentally reliving or pre-living events — is derived from average waypoint vividness.

Key capabilities:

• Two journey types: retrospection (reliving past) and prospection (imagining future)
• Vividness modeling: decays with temporal distance, floors at 0.1
• Confabulation: 10% random probability of a false waypoint detail
• State machine: forming -> traveling -> arrived -> reflecting -> complete
• Autonoetic labels: vivid / clear / hazy / fragmented / dreamlike

Add to your Gemfile:

gem 'lex-mental-time-travel'

Or install directly:

gem install lex-mental-time-travel

require 'legion/extensions/mental_time_travel'

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

# Quick retrospective journey
result = client.retrospect(
  destination: 'last deployment incident',
  waypoints: [
    { description: 'deploy triggered', distance: 3 },
    { description: 'error rate spiked', distance: 2 },
    { description: 'rollback executed', distance: 1 }
  ]
)

# Quick prospective journey
result = client.prospect(
  destination: 'next release cycle',
  waypoints: [
    { description: 'feature freeze', distance: 1 },
    { description: 'staging deploy', distance: 2 },
    { description: 'production release', distance: 3 }
  ]
)
journey_id = result[:journey][:id]

# Check for confabulated waypoints
client.confabulation_report(journey_id: journey_id)

# Check autonoetic level
client.autonoetic_status
# => { level: 0.72, label: :clear, journey_id: "..." }

# Stats
client.mental_time_travel_stats

bundle install
bundle exec rspec
bundle exec rubocop

MIT