Project Structure¶
src/probos/
├── __init__.py # Package root
├── __main__.py # Entry point (probos CLI)
├── config.py # Pydantic config models
├── runtime.py # Top-level orchestrator (2,762 lines, decomposed)
├── types.py # Core types (30+ dataclasses)
├── build_queue.py # Priority build queue
├── build_dispatcher.py # Automated builder dispatch
├── crew_profile.py # Crew identity + personality
├── sif.py # Structural Integrity Field
├── task_tracker.py # Agent task lifecycle
├── watch_rotation.py # Watch rotation + duty shifts
├── worktree_manager.py # Git worktree lifecycle
├── agents/ # Tool agents (deterministic) + department crews
│ ├── file_reader.py # read_file, stat_file
│ ├── file_writer.py # write_file (consensus-gated)
│ ├── directory_list.py # list_directory
│ ├── file_search.py # search_files
│ ├── shell_command.py # run_command (consensus-gated)
│ ├── http_fetch.py # http_fetch (rate-limited)
│ ├── introspect.py # explain_last, agent_info, system_health, why
│ ├── system_qa.py # Smoke tests for designed agents
│ ├── red_team.py # Independent verification + write checks
│ ├── corrupted.py # Test agent (deliberately wrong)
│ ├── utility/ # 10 bundled utility agents
│ │ ├── web_agents.py # WebSearch, PageReader, Weather, News
│ │ ├── language_agents.py # Translator, Summarizer
│ │ ├── productivity_agents.py # Calculator, Todo, NoteTaker
│ │ └── organizer_agents.py # Scheduler
│ ├── medical/ # Medical department (5 agents)
│ │ ├── diagnostician.py # Chief Medical Officer (Bones)
│ │ ├── vitals_monitor.py # Continuous health metrics (Chapel)
│ │ ├── surgeon.py # Targeted remediation (Chapel, dual-hatted)
│ │ ├── pharmacist.py # Configuration prescriptions (Keiko)
│ │ └── pathologist.py # Failure analysis (Cortez)
│ └── science/ # Science Analytical Pyramid (AD-560)
│ ├── data_analyst.py # Data Analyst (Kira)
│ ├── systems_analyst.py # Systems Analyst (Lynx)
│ └── research_specialist.py # Research Specialist (Atlas)
├── cognitive/ # LLM pipeline + self-modification + crew agents
│ ├── cognitive_agent.py # Instructions-first LLM agent base
│ ├── decomposer.py # NL → TaskDAG + DAG executor
│ ├── prompt_builder.py # Dynamic system prompt assembly
│ ├── llm_client.py # OpenAI-compatible + mock client
│ ├── working_memory.py # Bounded context assembly
│ ├── episodic.py # Episodic memory (Anchor Frames, ACT-R activation)
│ ├── attention.py # Priority scoring + focus tracking
│ ├── dreaming.py # 12-step dream consolidation
│ ├── dream_adapter.py # Dream cycle coordination
│ ├── workflow_cache.py # LRU pattern cache
│ ├── standing_orders.py # 4-tier instruction composition
│ ├── self_model.py # SystemSelfModel for grounding
│ ├── trust_dampening.py # Trust cascade dampening (AD-558)
│ ├── emergence_metrics.py # PID-based emergence measurement (AD-557)
│ ├── self_regulation.py # 3-tier cognitive self-regulation (AD-502–506)
│ ├── qualification_tests.py # Cognitive qualification probes
│ ├── domain_tests.py # Domain-specific qualification tests
│ ├── orientation.py # Agent orientation service
│ ├── architect.py # ArchitectAgent / First Officer (Meridian)
│ ├── builder.py # BuilderAgent / Chief Engineer
│ ├── code_reviewer.py # CodeReviewAgent
│ ├── counselor.py # CounselorAgent / Ship's Counselor (Echo)
│ ├── scout.py # ScoutAgent (Horizon)
│ ├── security_officer.py # SecurityAgent (Worf)
│ ├── operations_officer.py # OperationsAgent (O'Brien)
│ ├── engineering_officer.py # EngineeringAgent (LaForge)
│ ├── codebase_index.py # Codebase knowledge graph
│ ├── copilot_adapter.py # Visiting officer (Copilot SDK)
│ ├── agent_designer.py # LLM designs new agents from capability gaps
│ ├── self_mod.py # Self-modification pipeline
│ ├── code_validator.py # Static analysis for generated code
│ ├── sandbox.py # Isolated execution for untrusted agents
│ └── ... # + feedback, patcher, embeddings, proactive, etc.
├── cognitive_jit/ # Procedural Learning pipeline (AD-531–539)
│ ├── clustering.py # Episode clustering
│ ├── extraction.py # Procedure extraction
│ ├── store.py # Procedure store
│ ├── replay.py # Replay engine + fallback
│ ├── graduation.py # Dreyfus competency levels
│ ├── governance.py # Trust-gated promotion
│ ├── observational.py # Observational learning (Bandura)
│ ├── lifecycle.py # Decay, archival, dedup
│ └── gap_detection.py # Gap → qualification triggering
├── identity/ # W3C DID Identity (AD-441)
│ ├── did.py # DID generation + resolution
│ ├── credentials.py # Verifiable Credentials
│ ├── ledger.py # Identity Ledger (hash-chain)
│ └── birth_certificate.py # Agent + Ship birth certificates
├── ward_room/ # Agent Communication Fabric (AD-407–412)
│ ├── channels.py # Channel management (10 default)
│ ├── messages.py # Message storage + threading
│ ├── dm.py # Direct message channels
│ └── moderation.py # Content moderation + rate limiting
├── ships_records/ # Ship's Records (AD-434)
│ ├── notebooks.py # Agent notebook management
│ ├── duty_log.py # Duty log entries
│ └── captains_log.py # Captain's Log
├── startup/ # Runtime decomposition (AD-515–519)
│ ├── infrastructure.py # Phase 1: Core infrastructure
│ ├── structural_services.py # Phase 2: Structural services
│ ├── agent_fleet.py # Phase 3: Agent pool creation
│ ├── fleet_organization.py # Phase 4: Pool groups + departments
│ ├── cognitive_services.py # Phase 5: Skills, QA, self-mod
│ ├── communication.py # Phase 6: Channels + Discord
│ ├── dreaming.py # Phase 7: Dream engine setup
│ ├── results.py # Phase 8: Result persistence
│ ├── finalize.py # Phase 9: Final initialization
│ └── shutdown.py # Graceful shutdown sequence
├── routers/ # FastAPI routers (AD-515–519)
│ ├── agents.py # Agent management endpoints
│ ├── chat.py # Chat + intent processing
│ ├── wardroom.py # Ward Room API
│ ├── identity.py # DID + credential endpoints
│ ├── procedures.py # Cognitive JIT procedures
│ ├── records.py # Ship's Records API
│ ├── recreation.py # Recreation + games
│ └── ... # + 13 more domain routers
├── experience/ # User interface
│ ├── shell.py # Async REPL (42 slash commands)
│ ├── renderer.py # Real-time DAG execution display
│ ├── panels.py # Rich panel/table rendering
│ └── commands/ # Shell commands (AD-517)
│ ├── commands_status.py # /status, /agents, /ping, etc.
│ ├── commands_memory.py # /memory, /recall, /dream, etc.
│ ├── commands_knowledge.py # /knowledge, /search, /scout, etc.
│ ├── commands_directives.py # /orders, /directives, etc.
│ ├── commands_autonomous.py # /conn, /night-orders, /watch
│ ├── commands_procedure.py # /procedure, /gap, /qualify
│ └── ... # + 6 more command modules
├── recreation/ # Agent recreation system (AD-526)
│ ├── games.py # Game engine (tic-tac-toe, etc.)
│ └── creative.py # Creative expression channels
├── ontology/ # Vessel ontology (AD-513)
│ └── vessel.py # Crew manifest + cognitive grounding
├── storage/ # Abstract storage interfaces
│ └── connections.py # Cloud-ready DB connection layer
├── channels/ # Communication adapters
│ ├── base.py # Channel ABC
│ ├── discord_adapter.py # Discord integration
│ └── response_formatter.py # Format responses per channel
├── consensus/ # Multi-agent agreement
│ ├── quorum.py # Confidence-weighted voting
│ ├── trust.py # Bayesian Beta(α,β) reputation
│ ├── shapley.py # Shapley value attribution
│ └── escalation.py # 3-tier failure cascade
├── federation/ # Multi-node mesh
│ ├── bridge.py # ZeroMQ node bridge
│ ├── router.py # Intent forwarding + loop prevention
│ └── transport.py # Transport abstraction
├── knowledge/ # Persistent storage
│ ├── store.py # Git-backed operational state persistence
│ └── semantic.py # SemanticKnowledgeLayer (ChromaDB)
├── mesh/ # Agent coordination
│ ├── intent.py # Pub/sub bus with fan-out
│ ├── routing.py # Hebbian learning (SQLite)
│ ├── capability.py # Fuzzy matching registry
│ ├── gossip.py # SWIM-style state exchange
│ └── signal.py # TTL-enforced signals
├── substrate/ # Agent lifecycle
│ ├── agent.py # BaseAgent ABC (perceive/decide/act/report)
│ ├── registry.py # Async-safe agent index
│ ├── spawner.py # Template-based factory
│ ├── pool.py # Resource pools + health checks
│ ├── pool_group.py # PoolGroup + PoolGroupRegistry
│ ├── scaler.py # Demand-based pool scaling
│ ├── heartbeat.py # Periodic pulse loop
│ ├── event_log.py # Append-only SQLite audit log
│ ├── identity.py # Deterministic slot identity
│ └── skill_agent.py # SkillBasedAgent (dynamic skill dispatch)
└── utils/ # Shared utilities
config/
├── system.yaml # Main configuration
└── standing_orders/ # Constitution hierarchy
├── federation.md # Tier 1: Federation Constitution
├── ship.md # Tier 2: Ship Standing Orders
├── engineering.md # Tier 3: Department protocols
├── science.md
├── medical.md
├── security.md
├── operations.md
├── bridge.md
├── builder.md # Tier 4: Agent standing orders
├── architect.md
├── counselor.md
└── ... (15+ agent files)
ui/src/ # HXI — Human Experience Interface (React + Three.js)
├── canvas/ # WebGL cognitive mesh visualization
├── components/ # IntentSurface, MissionControl, SystemOrb, overlays
├── audio/ # TTS, speech input, sound engine
├── store/ # Zustand state management + TypeScript types
└── hooks/ # WebSocket connection to runtime
Wave 3 Decomposition (AD-515/516/517/518/519)¶
The three largest files in the codebase were decomposed into focused modules:
| Original File | Before | After | Reduction | New Package |
|---|---|---|---|---|
runtime.py |
5,321 lines | 2,762 lines | -48.1% | startup/ (10 modules) |
api.py |
3,109 lines | 295 lines | -90.5% | routers/ (21 modules) |
shell.py |
1,883 lines | 507 lines | -73.1% | experience/commands/ (13 modules) |