Skip to content

Phase 69: C2 Depth

Status: Complete Block: 8 (Consequence Enforcement & Scenario Expansion) Tests: 41 new (5 test files)

Summary

Phase 69 makes the C2 chain produce real effects: ATO limits air tempo, planning results influence AI decisions, deception injects phantom contacts, command hierarchy enforces authority, and burned zones reduce concealment.

What Was Built

69e: Burned Zone Concealment

  • _compute_terrain_modifiers() now queries incendiary_engine.get_burned_zones()
  • Units inside a burned zone have concealment reduced by zone.concealment_reduction
  • Multiple overlapping zones stack reductions (additive, floored at 0.0)
  • Gated by presence of incendiary_engine (implicitly by enable_fire_zones)

69a: ATO Sortie Consumption

  • ATOPlanningEngine.record_sortie(unit_id, end_time_s) increments sorties_today and sets last_sortie_end_time_s
  • ATOPlanningEngine.reset_daily_sorties(current_time_s) resets all aircraft sortie counts
  • get_state()/set_state() checkpoint the _aircraft dict
  • After air engagement routing in battle.py, record_sortie() is called for the attacker
  • In engine.py, daily reset at day boundary via int(time / 86400) comparison
  • Gated by enable_air_routing flag

69b: Planning Result Injection

  • PlanningProcessEngine.get_planning_result(unit_id) returns posture string if planning is COMPLETE
  • PlanningProcessEngine.consume_result(unit_id) returns and clears the result (one-shot)
  • Auto-generated COA: when complete_planning() is called and selected_coa is None, defaults to "ATTACK"
  • In DECIDE phase: consumed result boosts matching posture in school_adjustments by 0.10
  • Gated by enable_c2_friction flag

69c: Deception & FOW Injection

  • FogOfWarManager.deploy_decoy(), get_active_decoys(), update_decoys() passthrough to internal DeceptionEngine
  • Deception stratagem activation deploys deception_phantom_count (default 3) phantoms at random offsets near feint units
  • OBSERVE phase: active decoys inflate enemy_power in assessment (one phantom = one additional enemy unit)
  • Decoys degrade over time via update_decoys(dt) called each tick in the battle loop
  • Gated by enable_fog_of_war flag

69d: Command Hierarchy Enforcement

  • enable_command_hierarchy CalibrationSchema field (default False)
  • When enabled, scenario.py builds HierarchyTree from units_by_side: virtual HQ per side at DIVISION echelon, each unit as COMPANY child
  • TaskOrgManager wraps the hierarchy with default ORGANIC relationships
  • CommandEngine instantiated and passed to OrderPropagationEngine (replaces None)
  • Authority enforcement flow: propagate_order() → can_issue_order() → hierarchy check
  • Self-orders pass (unit in own CoC), cross-side orders fail, HQ-to-subordinate passes

New CalibrationSchema Fields (2)

Field Type Default Consumer
enable_command_hierarchy bool False scenario.py — CommandEngine instantiation
deception_phantom_count int 3 battle.py — phantoms per deception stratagem

Files Modified

File Changes
simulation/calibration.py +2 fields
simulation/battle.py 69a (sortie recording), 69b (planning result), 69c (decoy deploy + assessment), 69e (burned zone)
simulation/engine.py 69a (daily sortie reset)
simulation/scenario.py 69d (hierarchy building, CommandEngine instantiation, units_by_side param)
c2/orders/air_orders.py 69a (record_sortie, reset_daily_sorties, get_state/set_state)
c2/planning/process.py 69b (get_planning_result, consume_result, auto-COA)
detection/fog_of_war.py 69c (deploy_decoy, get_active_decoys, update_decoys passthroughs)

Test Files (5)

File Tests
test_phase_69a_ato_sortie.py 7
test_phase_69b_planning_result.py 8
test_phase_69c_deception_fow.py 11
test_phase_69d_command_hierarchy.py 8
test_phase_69e_burned_zone.py 7
Total 41

Lessons Learned

  • Turnaround check is < not <=: get_available_sorties() uses current_time - last_sortie_end_time < turnaround_time, so at exactly turnaround time the aircraft becomes available.
  • Auto-COA on complete: complete_planning() is the natural place to auto-generate a COA — it's the terminal transition and no external code calls it for the result before this point.
  • Deception is assessment inflation, not detection spoofing: Deploying decoys through FOW and inflating enemy_power in assessment is simpler and more effective than trying to inject phantom contacts into the detection loop.
  • Virtual HQ pattern scales: Creating per-side virtual HQ nodes at DIVISION echelon with all units as COMPANY children gives correct authority semantics without requiring org YAML parsing.
  • _create_engines() needed units_by_side param: Command hierarchy building requires knowledge of the force structure, which wasn't previously passed to the engine factory method.

Postmortem

Scope: On target

All 5 planned substeps delivered (69a–69e). No items dropped or deferred. One unplanned item added: update_decoys() tick-loop wiring (caught during postmortem integration audit).

Quality: High

  • 41 tests across 5 files covering unit behavior, backward compat, edge cases
  • No TODOs/FIXMEs in new code
  • All public methods have type hints and docstrings

Integration: Fully wired (after postmortem fix)

  • Postmortem caught that update_decoys(dt) was defined but never called in the tick loop — decoys would never degrade. Fixed by adding the call in battle.py after the fire zone damage block.
  • All other new methods verified called from simulation loop.
  • Structural test updated: consumer check now includes scenario.py; enable_command_hierarchy deferred from scenario exercise requirement.

Deficits: 0 new

  • No new known limitations. The plan's risk notes (69c "FOW.update() still not called in full cycle") were partially addressed — decoy degradation is wired but full FOW detection cycling remains a larger future effort.

Cross-doc audit: 3 issues found and fixed

  1. README badges stale (phase-68 → phase-69, test count 9,044 → 9,085)
  2. mkdocs.yml missing phase-69 devlog nav entry
  3. Test counts updated from 40 → 41 across all docs after postmortem added degradation test