d14150da75
Implement intelligent agent learning from Knowledge Graph execution history with per-task-type expertise tracking, recency bias, and learning curves. ## Phase 5.3 Implementation ### Learning Infrastructure (✅ Complete) - LearningProfileService with per-task-type expertise metrics - TaskTypeExpertise model tracking success_rate, confidence, learning curves - Recency bias weighting: recent 7 days weighted 3x higher (exponential decay) - Confidence scoring prevents overfitting: min(1.0, executions / 20) - Learning curves computed from daily execution windows ### Agent Scoring Service (✅ Complete) - Unified AgentScore combining SwarmCoordinator + learning profiles - Scoring formula: 0.3*base + 0.5*expertise + 0.2*confidence - Rank agents by combined score for intelligent assignment - Support for recency-biased scoring (recent_success_rate) - Methods: rank_agents, select_best, rank_agents_with_recency ### KG Integration (✅ Complete) - KGPersistence::get_executions_for_task_type() - query by agent + task type - KGPersistence::get_agent_executions() - all executions for agent - Coordinator::load_learning_profile_from_kg() - core KG→Learning integration - Coordinator::load_all_learning_profiles() - batch load for multiple agents - Convert PersistedExecution → ExecutionData for learning calculations ### Agent Assignment Integration (✅ Complete) - AgentCoordinator uses learning profiles for task assignment - extract_task_type() infers task type from title/description - assign_task() scores candidates using AgentScoringService - Fallback to load-based selection if no learning data available - Learning profiles stored in coordinator.learning_profiles RwLock ### Profile Adapter Enhancements (✅ Complete) - create_learning_profile() - initialize empty profiles - add_task_type_expertise() - set task-type expertise - update_profile_with_learning() - update swarm profiles from learning ## Files Modified ### vapora-knowledge-graph/src/persistence.rs (+30 lines) - get_executions_for_task_type(agent_id, task_type, limit) - get_agent_executions(agent_id, limit) ### vapora-agents/src/coordinator.rs (+100 lines) - load_learning_profile_from_kg() - core KG integration method - load_all_learning_profiles() - batch loading for agents - assign_task() already uses learning-based scoring via AgentScoringService ### Existing Complete Implementation - vapora-knowledge-graph/src/learning.rs - calculation functions - vapora-agents/src/learning_profile.rs - data structures and expertise - vapora-agents/src/scoring.rs - unified scoring service - vapora-agents/src/profile_adapter.rs - adapter methods ## Tests Passing - learning_profile: 7 tests ✅ - scoring: 5 tests ✅ - profile_adapter: 6 tests ✅ - coordinator: learning-specific tests ✅ ## Data Flow 1. Task arrives → AgentCoordinator::assign_task() 2. Extract task_type from description 3. Query KG for task-type executions (load_learning_profile_from_kg) 4. Calculate expertise with recency bias 5. Score candidates (SwarmCoordinator + learning) 6. Assign to top-scored agent 7. Execution result → KG → Update learning profiles ## Key Design Decisions ✅ Recency bias: 7-day half-life with 3x weight for recent performance ✅ Confidence scoring: min(1.0, total_executions / 20) prevents overfitting ✅ Hierarchical scoring: 30% base load, 50% expertise, 20% confidence ✅ KG query limit: 100 recent executions per task-type for performance ✅ Async loading: load_learning_profile_from_kg supports concurrent loads ## Next: Phase 5.4 - Cost Optimization Ready to implement budget enforcement and cost-aware provider selection.
88 lines
1.9 KiB
TOML
88 lines
1.9 KiB
TOML
# Multi-IA Router Configuration
|
|
# Phase 0: Configuration for LLM provider selection
|
|
|
|
[routing]
|
|
# Default provider if no specific routing rules match
|
|
default_provider = "claude"
|
|
|
|
# Enable cost tracking
|
|
cost_tracking_enabled = true
|
|
|
|
# Enable fallback on provider failure
|
|
fallback_enabled = true
|
|
|
|
[providers.claude]
|
|
enabled = true
|
|
# ANTHROPIC_API_KEY environment variable required
|
|
api_key = "${ANTHROPIC_API_KEY}"
|
|
model = "claude-sonnet-4-5-20250929"
|
|
max_tokens = 8192
|
|
temperature = 0.7
|
|
|
|
# Cost per 1M tokens (input/output)
|
|
cost_per_1m_input = 3.00
|
|
cost_per_1m_output = 15.00
|
|
|
|
[providers.openai]
|
|
enabled = true
|
|
# OPENAI_API_KEY environment variable required
|
|
api_key = "${OPENAI_API_KEY}"
|
|
model = "gpt-4o"
|
|
max_tokens = 4096
|
|
temperature = 0.7
|
|
|
|
# Cost per 1M tokens (input/output)
|
|
cost_per_1m_input = 2.50
|
|
cost_per_1m_output = 10.00
|
|
|
|
[providers.gemini]
|
|
enabled = true
|
|
# GOOGLE_API_KEY environment variable required
|
|
api_key = "${GOOGLE_API_KEY}"
|
|
model = "gemini-2.0-flash"
|
|
max_tokens = 8192
|
|
temperature = 0.7
|
|
|
|
# Cost per 1M tokens (input/output)
|
|
cost_per_1m_input = 0.30
|
|
cost_per_1m_output = 1.20
|
|
|
|
[providers.ollama]
|
|
enabled = true
|
|
# Local Ollama instance, no API key needed
|
|
url = "${OLLAMA_URL:-http://localhost:11434}"
|
|
model = "llama3.2"
|
|
max_tokens = 4096
|
|
temperature = 0.7
|
|
|
|
# No cost for local models
|
|
cost_per_1m_input = 0.00
|
|
cost_per_1m_output = 0.00
|
|
|
|
# Routing rules: assign providers based on task characteristics
|
|
[[routing_rules]]
|
|
name = "architecture_design"
|
|
condition = { task_type = "architecture" }
|
|
provider = "claude"
|
|
model_override = "claude-opus-4-20250514"
|
|
|
|
[[routing_rules]]
|
|
name = "code_generation"
|
|
condition = { task_type = "development" }
|
|
provider = "claude"
|
|
|
|
[[routing_rules]]
|
|
name = "documentation"
|
|
condition = { task_type = "documentation" }
|
|
provider = "openai"
|
|
|
|
[[routing_rules]]
|
|
name = "monitoring"
|
|
condition = { task_type = "monitoring" }
|
|
provider = "gemini"
|
|
|
|
[[routing_rules]]
|
|
name = "local_testing"
|
|
condition = { environment = "development" }
|
|
provider = "ollama"
|