Module: Legion::Extensions::Apollo::Runners::Maintenance

Includes:
Helpers::Lex
Included in:
Client
Defined in:
lib/legion/extensions/apollo/runners/maintenance.rb

Instance Method Summary collapse

Instance Method Details

#archive_stale(days: 90) ⇒ Object 



16
17
18
 
# File 'lib/legion/extensions/apollo/runners/maintenance.rb', line 16

def archive_stale(days: 90, **)
  { action: :archive_stale, days: days }
end

#check_corroboration ⇒ Object

rubocop:disable Metrics/CyclomaticComplexity



57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
 
# File 'lib/legion/extensions/apollo/runners/maintenance.rb', line 57

def check_corroboration(**) # rubocop:disable Metrics/CyclomaticComplexity
  return { success: false, error: 'apollo_data_not_available' } unless defined?(Legion::Data::Model::ApolloEntry)

  candidates = Legion::Data::Model::ApolloEntry.where(status: 'candidate').exclude(embedding: nil).all
  confirmed = Legion::Data::Model::ApolloEntry.where(status: 'confirmed').exclude(embedding: nil).all

  promoted = 0

  candidates.each do |candidate|
    match = confirmed.find do |conf|
      next unless conf.content_type == candidate.content_type

      sim = Helpers::Similarity.cosine_similarity(
        vec_a: candidate.embedding, vec_b: conf.embedding
      )
      Helpers::Similarity.above_corroboration_threshold?(similarity: sim)
    end

    next unless match

    candidate_provider = candidate.respond_to?(:source_provider) ? candidate.source_provider : nil
    match_provider     = match.respond_to?(:source_provider) ? match.source_provider : nil
    both_known = known_provider?(candidate_provider) && known_provider?(match_provider)
    next if both_known && candidate_provider == match_provider

    # Also reject if same source_channel (same data pipeline)
    candidate_channel = candidate.respond_to?(:source_channel) ? candidate.source_channel : nil
    match_channel = match.respond_to?(:source_channel) ? match.source_channel : nil
    next if candidate_channel && match_channel && candidate_channel == match_channel

    candidate.update(
      status:       'confirmed',
      confirmed_at: Time.now,
      confidence:   Helpers::Confidence.apply_corroboration_boost(confidence: candidate.confidence),
      updated_at:   Time.now
    )

    Legion::Data::Model::ApolloRelation.create(
      from_entry_id: candidate.id,
      to_entry_id:   match.id,
      relation_type: 'similar_to',
      source_agent:  'system:corroboration',
      weight:        1.0
    )

    promoted += 1
  end

  { success: true, promoted: promoted, scanned: candidates.size }
rescue Sequel::Error => e
  { success: false, error: e.message }
end

#force_decay(factor: 0.5) ⇒ Object 



12
13
14
 
# File 'lib/legion/extensions/apollo/runners/maintenance.rb', line 12

def force_decay(factor: 0.5, **)
  { action: :force_decay, factor: factor }
end

#resolve_dispute(entry_id:, resolution:) ⇒ Object 



20
21
22
 
# File 'lib/legion/extensions/apollo/runners/maintenance.rb', line 20

def resolve_dispute(entry_id:, resolution:, **)
  { action: :resolve_dispute, entry_id: entry_id, resolution: resolution }
end

#run_decay_cycle(alpha: nil, min_confidence: nil) ⇒ Object 



24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
 
# File 'lib/legion/extensions/apollo/runners/maintenance.rb', line 24

def run_decay_cycle(alpha: nil, min_confidence: nil, **)
  alpha ||= decay_alpha
  min_confidence ||= decay_threshold

  return { decayed: 0, archived: 0 } unless defined?(Legion::Data) && Legion::Data.respond_to?(:connection) && Legion::Data.connection

  conn = Legion::Data.connection

  # Power-law: per-cycle decay factor decreases as entries age
  # Factor = (hours_old / (hours_old + 1)) ^ alpha
  # Recent entries (small hours_old) get a factor closer to 0 (more decay)
  # Old entries (large hours_old) get a factor closer to 1 (less decay)
  hours_expr = Sequel.lit(
    'GREATEST(EXTRACT(EPOCH FROM (NOW() - COALESCE(updated_at, created_at))) / 3600.0, 1.0)'
  )
  decay_factor = Sequel.lit(
    'POWER(CAST(? AS double precision) / (CAST(? AS double precision) + 1.0), ?)', hours_expr, hours_expr, alpha
  )

  decayed = conn[:apollo_entries]
            .exclude(status: 'archived')
            .update(confidence: Sequel[:confidence] * decay_factor)

  archived = conn[:apollo_entries]
             .where { confidence < min_confidence }
             .exclude(status: 'archived')
             .update(status: 'archived')

  { decayed: decayed, archived: archived, alpha: alpha, threshold: min_confidence }
rescue Sequel::Error => e
  { decayed: 0, archived: 0, error: e.message }
end