Development Log

Rolling record of implementation decisions, changes, and lessons learned across all phases of Stochastic Warfare development.

Phases

Phase	Focus	Status	Log
0	Project Scaffolding	Complete	phase-0.md
1	Terrain & Environment Foundation	Complete	phase-1.md
2	Entity System & Movement	Complete	phase-2.md
3	Detection & Intelligence	Complete	phase-3.md
4	Combat Resolution & Morale	Complete	phase-4.md
5	C2 Infrastructure	Complete	phase-5.md
6	Logistics & Supply	Complete	phase-6.md
7	Engagement Validation	Complete	phase-7.md
8	AI & Planning	Complete	phase-8.md
9	Simulation Orchestration	Complete	phase-9.md
10	Campaign Validation	Complete	phase-10.md
11	Core Fidelity Fixes	Complete	phase-11.md
12	Deep Systems Rework	Complete	phase-12.md
13	Performance Optimization	Complete	phase-13.md
14	Tooling & Developer Experience	Complete	phase-14.md
15	Real-World Terrain & Data Pipeline	Complete	phase-15.md
16	Electronic Warfare	Complete	phase-16.md
17	Space & Satellite Domain	Complete	phase-17.md
18	NBC/CBRN Effects	Complete	phase-18.md
19	Doctrinal AI Schools	Complete	phase-19.md
20	WW2 Era	Complete	phase-20.md
21	WW1 Era	Complete	phase-21.md
22	Napoleonic Era	Complete	phase-22.md
23	Ancient & Medieval Era	Complete	phase-23.md
24	Unconventional & Prohibited Warfare	Complete	phase-24.md
25	Engine Wiring & Integration Sprint	Complete	phase-25.md
26	Core Polish & Configuration	Complete	phase-26.md
27	Combat System Completeness	Complete	phase-27.md
28	Modern Era Data Package	Complete	phase-28.md
28.5	Directed Energy Weapons	Complete	phase-28.5.md
29	Historical Era Data Expansion	Complete	phase-29.md
30	Scenario & Campaign Library	Complete	phase-30.md
31	Documentation Site (GitHub Pages)	Complete	phase-31.md
32	API & Service Foundation	Complete	phase-32.md
33	Frontend Foundation & Scenario Browser	Complete	phase-33.md
34	Run Results & Analysis Dashboard	Complete	phase-34.md
35	Tactical Map & Spatial Visualization	Complete	phase-35.md
36	Scenario Tweaker & Polish	Complete	phase-36.md
37	Integration Fixes & E2E Validation	Complete	phase-37.md
38	Map & Chart Enhancements	Complete	phase-38.md
39	Quality, Performance & Packaging	Complete	phase-39.md
40	Battle Loop Foundation	Complete	phase-40.md
41	Combat Depth	Complete	phase-41.md
42	Tactical Behavior	Complete	phase-42.md
43	Domain-Specific Resolution	Complete	phase-43.md
44	Environmental & Subsystem Integration	Complete	phase-44.md
45	Mathematical Model Audit & Hardening	Complete	phase-45.md
46	Scenario Data Cleanup & Expansion	Complete	phase-46.md
47	Full Recalibration & Validation	Complete	phase-47.md
48	Block 5 Deficit Resolution	Complete	phase-48.md
49	Calibration Schema Hardening	Complete	phase-49.md
50	Combat Fidelity Polish	Complete	phase-50.md
51	Naval Combat Completeness	Complete	phase-51.md
52	Environmental Continuity	Complete	phase-52.md
53	C2 & AI Completeness	Complete	phase-53.md
54	Era-Specific & Domain Sub-Engine Wiring	Complete	phase-54.md
55	Resolution & Scenario Migration	Complete	phase-55.md
56	Performance & Logistics	Complete	phase-56.md
57	Full Validation & Regression	Complete	phase-57.md
58	Structural Verification & Core Combat Wiring	Complete	phase-58.md
59	Atmospheric & Ground Environment Wiring	Complete	phase-59.md
60	Obscurants, Fire, & Visual Environment	Complete	phase-60.md
61	Maritime, Acoustic, & EM Environment	Complete	phase-61.md
62	Human Factors, CBRN, & Air Combat Environment	Complete	phase-62.md
63	Cross-Module Feedback Loops	Complete	phase-63.md
64	C2 Friction & Command Delay	Complete	phase-64.md
65	Space & EW Sub-Engine Activation	Complete	phase-65.md
66	Unconventional, Naval, & Cleanup	Complete	phase-66.md
67	Integration Validation & Recalibration	Complete	phase-67.md
Block 8: Consequence Enforcement & Scenario Expansion
68	Consequence Enforcement	Complete	phase-68.md
69	C2 Depth	Complete	phase-69.md
70	Performance Optimization	Complete	phase-70.md
71	Missile & Carrier Ops Completion	Complete	phase-71.md
72	Checkpoint & State Completeness	Complete	phase-72.md
73	Historical Scenario Correctness	Complete	phase-73.md
74	Combat Engine Unit Tests	Complete	phase-74.md
75	Simulation Core & Domain Unit Tests	Complete	phase-75.md
76	API Robustness	Complete	phase-76.md
77	Frontend Accessibility	Complete	phase-77.md
78	P2 Environment Wiring	Complete	phase-78.md
79	CI/CD & Packaging	Complete	phase-79.md
80	API & Frontend Sync	Complete	phase-80.md
81	Recalibration & Validation	Complete	phase-81.md
82	Block 8 Postmortem & Documentation	Complete	phase-82.md
Block 9: Performance at Scale
83	Profiling Infrastructure	Complete	phase-83.md
84	Spatial Culling & Scan Scheduling	Complete	phase-84.md
85	LOD & Aggregation	Complete	phase-85.md
86	Engagement & Calibration Optimization	Complete	phase-86.md
87	Expanded Numba JIT	Complete	phase-87.md
88	SoA Data Layer	Complete	phase-88.md
89	Per-Side Parallelism	Complete	phase-89.md
90	Validation & Benchmarking	Complete	phase-90.md
91	Scenario Recalibration & Regression	Complete	phase-91.md
Block 10: UI Depth & Engine Exposure
92	API Analytics & Frame Enrichment	Complete	phase-92.md
93	Results Dashboard Depth	Complete	phase-93.md
94	Tactical Map Enrichment	Complete	phase-94.md
95	Calibration & Scenario Editor Depth	Complete	phase-95.md
96	Analysis & Event Interaction	Complete	phase-96.md
97	Data Catalog & Block 10 Validation	Complete	phase-97.md

Post-MVP Refinement Index

Known limitations and deferred improvements logged during implementation. Review these after MVP is functional.

Phase	Item	Section
0	Checkpoint format longevity (pickle fragility) (resolved Phase 57 — JSON serialization with NumpyEncoder, legacy pickle fallback)	phase-0.md — Open Questions
3	Track-to-target association needs nearest-neighbor gating (resolved Phase 11b)	phase-3.md — Known Limitations
3	Environment data threading (caller responsibility) (accepted limitation — explicit coupling is simpler than implicit dependency injection)	phase-3.md — Known Limitations
3	Passive sonar bearing is placeholder (random, not geometric) (resolved Phase 11b)	phase-3.md — Known Limitations
3	No sensor FOV filtering against observer heading (resolved Phase 11b)	phase-3.md — Known Limitations
3	Single-scan detection (no dwell/integration gain) (resolved Phase 11b)	phase-3.md — Known Limitations
3	Test coverage gap (296→334 vs planned 455; backfilled C2-facing APIs) (resolved — 8,000+ tests across 57 phases massively exceed original target)	phase-3.md — Known Limitations
4	Ballistic trajectory uses simplified drag (no Mach-dependent Cd) (resolved Phase 11a)	phase-4.md — Known Limitations
4	DeMarre penetration approximation (no obliquity, composite, reactive armor) (resolved Phase 11a)	phase-4.md — Known Limitations
4	HEAT penetration is range-independent (accepted limitation — physically correct for shaped charge jets)	phase-4.md — Known Limitations
4	Submarine evasion simplified probability model (resolved Phase 12c)	phase-4.md — Known Limitations
4	Mine trigger model lacks detailed ship signature interaction (resolved Phase 12c)	phase-4.md — Known Limitations
4	Carrier ops deck management abstracted (no individual spot tracking) (accepted limitation — below simulation granularity)	phase-4.md — Known Limitations
4	Morale Markov is discrete-time (not continuous-time) (resolved Phase 12d)	phase-4.md — Known Limitations
4	PSYOP model is simplified effectiveness roll (resolved Phase 12d)	phase-4.md — Known Limitations
4	Naval damage control lacks compartment flooding model (resolved Phase 12c)	phase-4.md — Known Limitations
4	Air combat lacks detailed flight dynamics / energy-maneuverability (resolved Phase 12c)	phase-4.md — Known Limitations
4	Environment→combat coupling partial: air_combat, air_defense, naval_surface, indirect_fire lack env coupling (resolved Phase 11a)	phase-4.md — Known Limitations
5	No multi-hop propagation (single hop issuer→recipient only) (resolved Phase 12a)	phase-5.md — Known Limitations
5	No terrain-based LOS check for communications (resolved Phase 52c — diffraction model + exempt types + has_los bug fix)	phase-5.md — Known Limitations
5	Simplified FSCL (east-west line, not arbitrary polyline) (resolved Phase 12a)	phase-5.md — Known Limitations
5	No ATO planning cycle (structures only, generation deferred to Phase 9/Future) (resolved Phase 12a)	phase-5.md — Known Limitations
5	No JTAC/FAC observer model for CAS (resolved Phase 12a)	phase-5.md — Known Limitations
5	Messenger comm type has no terrain traversal or intercept risk (accepted limitation — courier engine handles Napoleonic era; modern MESSENGER is niche)	phase-5.md — Known Limitations
6	No supply optimization solver (pull-based nearest depot only) (resolved Phase 12b)	phase-6.md — Known Limitations
6	No multi-echelon supply chain (direct depot-to-unit) (resolved Phase 12b)	phase-6.md — Known Limitations
6	Simplified transport vulnerability (no escort effects) (resolved Phase 12b)	phase-6.md — Known Limitations
6	Medical M/M/c queueing approximate (exponential service) (resolved Phase 12b)	phase-6.md — Known Limitations
6	Engineering times deterministic (no stochastic variation) (resolved Phase 11c)	phase-6.md — Known Limitations
6	No fuel gating on movement (tracked but not enforced) (resolved Phase 11c)	phase-6.md — Known Limitations
6	Blockade effectiveness simplified (flat per-ship probability) (resolved Phase 51d — DisruptionEngine wired, blockade queries in campaign.py)	phase-6.md — Known Limitations
6	Captured supply efficiency flat 50% (no compatibility check) (dormant capability — implemented and tested but no game mechanic triggers depot capture; reserved for future campaign planning work)	phase-6.md — Known Limitations
6	No local water procurement (always from rear depots) (accepted limitation — below simulation granularity)	phase-6.md — Known Limitations
6	No ammunition production (scenario-defined depots only) (accepted limitation — scenarios define initial supply levels)	phase-6.md — Known Limitations
6	VLS non-reloadable-at-sea enforcement deferred to naval combat integration (resolved Phase 56d — exhaustion enforced, port reload deferred)	phase-6.md — Known Limitations
7	73 Easting exchange_ratio = inf (detection asymmetry prevents blue losses) (resolved Phase 30 — calibration tuning: visibility 800m, red engagement 1500m, thermal_contrast 1.5)	phase-7.md — Known Limitations
7	No fire rate limiting (units fire once per tick regardless of ROF) (resolved Phase 11a)	phase-7.md — Known Limitations
7	Uniform target_size_modifier (applies to both sides equally) (resolved Phase 11a)	phase-7.md — Known Limitations
7	No wave attack modeling (all units advance simultaneously) (resolved Phase 11c)	phase-7.md — Known Limitations
7	Pre-scripted behavior only (no tactical AI, deferred to Phase 8) (resolved Phase 8 — OODA FSM, commander personalities, doctrine templates, MDMP)	phase-7.md — Known Limitations
7	Falklands simplified — Sheffield attack only, no San Carlos raids (resolved Phase 30 — added San Carlos air raids + Goose Green ground scenarios)	phase-7.md — Known Limitations
7	Synthetic terrain (programmatic heightmaps, not real topographic data) (resolved Phase 15 — real-world terrain pipeline)	phase-7.md — Known Limitations
7	No logistics in validation scenarios (short engagements) (accepted limitation — deliberately short engagements)	phase-7.md — Known Limitations
7	No C2 propagation in validation (direct behavior, no order chain) (accepted limitation — deliberately short engagements)	phase-7.md — Known Limitations
7	Simplified force compositions (representative samples, not complete OOB) (resolved Phase 30 — expanded 73 Easting OOB, diverse modern/historical scenarios)	phase-7.md — Known Limitations
8	Named doctrinal schools (Clausewitzian AI, Sun Tzu AI) deferred to Future Phases (resolved Phase 19)	phase-8.md — Known Limitations
8	COA wargaming is analytical (Lanchester), not full nested simulation (accepted limitation — Lanchester captures key force-ratio dynamics)	phase-8.md — Known Limitations
8	No terrain-specific COA generation (e.g., no river crossing planning detail) (accepted limitation — COA operates at force-ratio level)	phase-8.md — Known Limitations
8	Implied task tables are simplified (not full FM 5-0 comprehensive list) (accepted limitation — captures key task categories)	phase-8.md — Known Limitations
8	No multi-echelon simultaneous planning (each commander plans independently) (accepted limitation — by design; avoids coordination complexity)	phase-8.md — Known Limitations
8	Estimates update periodically, not reactively to every event (accepted limitation — event-driven updates not needed at current tick resolution)	phase-8.md — Known Limitations
8	Stratagems are opportunity-evaluated, not proactively planned in COA (accepted limitation — by design; see Phase 53c wiring)	phase-8.md — Known Limitations
8	Brigade echelon decision hardcodes echelon_level=9 in result (cosmetic) (resolved Phase 11d)	phase-8.md — Known Limitations
9	No force aggregation/disaggregation — all units at individual resolution (resolved Phase 13 postmortem)	phase-9.md — Known Limitations
9	Single-threaded simulation loop (required for deterministic PRNG replay) (accepted limitation — required for deterministic reproducibility)	phase-9.md — Known Limitations
9	No auto-resolve — every engagement runs full tactical resolution (resolved Phase 13a-6)	phase-9.md — Known Limitations
9	Simplified strategic movement (no detailed operational pathfinding) (accepted limitation — campaign-scale abstraction)	phase-9.md — Known Limitations
9	Fixed reinforcement schedule (no Poisson/stochastic arrivals) (accepted limitation — scenario-defined determinism)	phase-9.md — Known Limitations
9	No naval campaign management (structurally supported but untested) (resolved — 6 naval scenarios run through CampaignManager with 23+ dedicated tests)	phase-9.md — Known Limitations
9	Synthetic terrain only (programmatic heightmaps, not real topographic data) (resolved Phase 15 — real-world terrain pipeline)	phase-9.md — Known Limitations
9	LOS cache is per-tick only (cleared each tick after movement) (resolved Phase 13 postmortem — selective invalidation wired)	phase-9.md — Known Limitations
9	No weather evolution mid-campaign beyond WeatherEngine.step() (resolved — WeatherEngine.update() implements full Markov transitions, diurnal temperature, O-U wind; called every tick)	phase-9.md — Known Limitations
9	Viewshed vectorization deferred (lower priority) (resolved Phase 13a-5)	phase-9.md — Known Limitations
9	STRtree for infrastructure spatial queries still deferred (resolved Phase 13a-2)	phase-9.md — Known Limitations
10	No fire rate limiting — units fire once per tick regardless of ROF (inherited) (resolved Phase 11a — fire rate cooldown)	phase-10.md — Known Limitations
10	No wave attack modeling — all red units advance simultaneously (inherited) (resolved Phase 11c — wave attack mechanics)	phase-10.md — Known Limitations
10	Campaign AI decisions coarse — OODA at echelon timing, may not produce tactical posture changes in short runs (accepted limitation — OODA timing is by design)	phase-10.md — Known Limitations
10	Simplified force compositions — representative samples, not complete historical OOB (resolved Phase 30 — expanded OOBs, diverse modern/historical scenarios)	phase-10.md — Known Limitations
10	Synthetic terrain — programmatic heightmaps, not real topographic data (resolved Phase 15 — real-world terrain pipeline)	phase-10.md — Known Limitations
9/10	(duplicate of Phase 9 fixed reinforcement entry above — consolidated)
10	No force aggregation/disaggregation — all units individually tracked (resolved Phase 13 postmortem)	phase-10.md — Known Limitations
10	AI expectation matching approximate — string-based posture detection (accepted limitation — sufficient for posture change validation)	phase-10.md — Known Limitations
10	Campaign metrics proxy territory control via survival fraction not spatial (accepted limitation — survival fraction captures campaign outcomes)	phase-10.md — Known Limitations
11	Fuel gating not wired to stockpile in battle.py (resolved Phase 12b)	phase-11.md — Known Limitations
11	Wave assignments are manual (no AI auto-assignment) (accepted limitation — explicit is better than implicit)	phase-11.md — Known Limitations
11	Integration gain caps at 4 scans (resolved Phase 26c — configurable max_integration_scans)	phase-11.md — Known Limitations
11	Armor type YAML data missing (resolved Phase 26c — armor_type field + 6 YAML files)	phase-11.md — Known Limitations
16	EW engines not yet wired into simulation engine tick loop (resolved Phase 25c)	phase-16.md — Known Limitations
16	No DRFM detailed waveform modeling (simplified effectiveness parameter) (accepted limitation — captures key jamming effects without specialist depth)	phase-16.md — Known Limitations
16	TDOA geolocation uses simplified centroid-shift algorithm (accepted limitation — captures geolocation accuracy without DSP complexity)	phase-16.md — Known Limitations
16	No cooperative jamming between multiple platforms (accepted limitation — single-jammer model captures key dynamics)	phase-16.md — Known Limitations
16	Campaign-level EW validation deferred (component-level only) (resolved — EW wired into engine tick loop Phase 25c; 4 scenarios with ew_config exercise campaign-level EW)	phase-16.md — Known Limitations
17	Simplified Keplerian orbits (no SGP4/TLE, no atmospheric drag for LEO decay) (accepted limitation — captures pass timing and coverage gaps)	phase-17.md — Known Limitations
17	No detailed satellite bus modeling (power, thermal, attitude control) (accepted limitation — below simulation granularity)	phase-17.md — Known Limitations
17	No space-based SIGINT integration with Phase 16 SIGINT engine (resolved Phase 52d — inverse-variance weighted fusion in engine.py tick loop)	phase-17.md — Known Limitations
17	Debris cascade model is statistical (no individual fragment tracking) (accepted limitation — below simulation granularity)	phase-17.md — Known Limitations
17	No satellite maneuvering or station-keeping fuel limits (accepted limitation — captures orbital mechanics without propulsion modeling)	phase-17.md — Known Limitations
17	No space weather effects (solar flares, radiation belt variations) (accepted limitation — below simulation granularity)	phase-17.md — Known Limitations
17	EMEnvironment GPS accuracy is not per-side (uses worst-case aggregation) (accepted limitation — worst-case aggregation is conservative)	phase-17.md — Postmortem
16/17	ScenarioLoader doesn't auto-wire EWEngine or SpaceEngine from YAML (resolved Phase 25a)	phase-17.md — Postmortem
16/17/18	ScenarioLoader doesn't auto-wire CBRNEngine from YAML (resolved Phase 25a)	phase-18.md — Postmortem
18	mopp_speed_factor parameter exists in movement engine but never passed from battle loop (resolved Phase 25c)	phase-18.md — Postmortem
18	Hardcoded terrain channeling thresholds in dispersal (5m valley/ridge, 50m offset) (resolved Phase 26b — DispersalConfig fields)	phase-18.md — Postmortem
18	Hardcoded fallback weather defaults in CBRN engine (wind=2.0, temp=20°C, cloud=0.5) (resolved Phase 26b — CBRNConfig fields)	phase-18.md — Postmortem
18	No automatic puff aging/cleanup mechanism in dispersal engine (resolved Phase 26c — cleanup_aged_puffs() + max_puff_age_s)	phase-18.md — Postmortem
19	CommanderEngine not wired into SimulationContext (resolved Phase 25d)	phase-19.md — Known Limitations
19	battle.py passes assessment=None to decide() (resolved Phase 25b)	phase-19.md — Known Limitations
16/17/18/19	ScenarioLoader doesn't auto-wire SchoolRegistry from YAML (resolved Phase 25a)	phase-19.md — Known Limitations
19	get_coa_score_weight_overrides() hook not called in battle loop (resolved Phase 25b) / get_stratagem_affinity() still deferred (resolved Phase 53c — called in DECIDE phase with school affinity weights)	phase-19.md — Postmortem
19	CommanderPersonality.school_id field defined but never read — schools assigned via SchoolRegistry instead (resolved Phase 53c — auto-assigns via SchoolRegistry in _apply_commander_assignments())	phase-19.md — Postmortem
20	Convoy engine does not model individual escort positions (abstract effectiveness parameter) (accepted limitation — campaign-scale abstraction)	phase-20.md — Known Limitations
20	Strategic bombing target regeneration is linear (no industrial interdependency graph) (accepted limitation — captures key targeting dynamics)	phase-20.md — Known Limitations
20	Fighter escort in strategic bombing is probability modifier, not full air combat sub-simulation (accepted limitation — campaign-scale abstraction)	phase-20.md — Known Limitations
16/17/18/19/20	ScenarioLoader doesn't auto-wire era-specific engines from YAML (resolved Phase 25a)	phase-20.md — Postmortem
21	Barrage drift is purely random walk — no systematic correction for observer feedback (resolved Phase 27d — observer correction)	phase-21.md — Known Limitations
21	Gas warfare engine does not model gas mask don time delay (units gain instant protection) (resolved Phase 27d — compute_exposure_during_don + get_effective_mopp_level)	phase-21.md — Known Limitations
21	Trench system has no wire-cutting mechanic (wire is a query attribute only, not a movement blocker) (accepted limitation — wire as defensive query attribute is sufficient)	phase-21.md — Known Limitations
16/17/18/19/20/21	ScenarioLoader doesn't auto-wire WW1 engines from YAML (resolved Phase 25a)	phase-21.md — Known Limitations
21	Gas warfare wind direction tolerance (60°) hardcoded — should be configurable in GasWarfareConfig (resolved Phase 26b — GasWarfareConfig.max_wind_angle_deg)	phase-21.md — Postmortem
21	Barrage/gas engines use hardcoded fallback RNG seed (42) when no RNG injected (resolved Phase 26a — rng required on all 23 engines)	phase-21.md — Postmortem
16/17/18/19/20/21/22	ScenarioLoader doesn't auto-wire Napoleonic engines from YAML (resolved Phase 25a)	phase-22.md — Known Limitations
22	Cavalry charge ignores terrain effects (speed not modified by slope or obstacles) (resolved Phase 27d — compute_cavalry_terrain_modifier)	phase-22.md — Known Limitations
22	No frontage/depth in melee — simplified to force ratio × formation modifier (resolved Phase 27d — compute_frontage_constraint)	phase-22.md — Known Limitations
22	Foraging ambush casualty rate hardcoded at 10% — should be configurable in ForagingConfig (resolved Phase 26b — ForagingConfig.ambush_casualty_rate)	phase-22.md — Known Limitations
22	Volley/melee/cavalry/courier/foraging engines use hardcoded fallback RNG seed (42) when no RNG injected (resolved Phase 26a — rng required on all 23 engines)	phase-22.md — Known Limitations
16/17/18/19/20/21/22/23	ScenarioLoader doesn't auto-wire Ancient/Medieval engines from YAML (resolved Phase 25a)	phase-23.md — Postmortem
28.5	DEWEngagementEvent has zero subscribers in production code (events published but never consumed) (closed Phase 37 — SimulationRecorder subscribes to base Event)	phase-28.5.md — Postmortem
28.5	dew_engine not used in simulation/engine.py or simulation/battle.py tick loops (wired in scenario only) (resolved Phase 37 — battle.py routes DEW via route_engagement)	phase-28.5.md — Postmortem
28.5	No scenario YAML references dew_config (engine can be created but no scenarios exercise it) (resolved — taiwan_strait scenario uses dew_config)	phase-28.5.md — Postmortem
28.5	ADUnitType.DEW not handled in air defense engagement logic (enum exists but no routing) (accepted limitation — DEW routes via weapon category, functionally complete)	phase-28.5.md — Postmortem
28.5	route_engagement() not called from battle.py (uses execute_engagement directly) — DEW routing untested in loop (resolved Phase 37 — battle.py calls route_engagement for all engagements)	phase-28.5.md — Postmortem
32	config_overrides accepted and stored but not applied to engine (calibration_overrides not injected before ScenarioLoader.load) (resolved Phase 37 — deep merge before load)	phase-32.md — Postmortem
32	GET /api/meta/terrain-types returns hardcoded list instead of deriving from TerrainConfig or data (resolved Phase 39d — uses LandCover enum)	phase-32.md — Postmortem
34	useBatchProgress hook has no dedicated test file (resolved Phase 39a — 4 dedicated tests)	phase-34.md — Postmortem
34	RunDetailPage tests don't cover cancelled/error run states (resolved Phase 39a — 2 tests)	phase-34.md — Postmortem
34	Analysis API responses untyped (Record<string, unknown>) — compare/sweep return free-form dicts (resolved Phase 39a — typed CompareResult/SweepResult)	phase-34.md — Postmortem
34	Hardcoded morale state names and event type strings in eventProcessing.ts (resolved Phase 39a — exported constants ENGAGEMENT_EVENTS, MORALE_EVENTS)	phase-34.md — Postmortem
34	Force time series reconstruction assumes no reinforcements (only counts destructions) (resolved Phase 37 — handles ReinforcementArrivedEvent)	phase-34.md — Postmortem
35	Frame data uses compact keys in storage but API expands to full names — slight redundancy (accepted limitation — cosmetic, no impact on functionality)	phase-35.md — Known Limitations
35	terrain_json and frames_json stored as TEXT blobs — no indexing, large runs may produce 2MB+ frames (accepted limitation — sufficient for current scenario sizes)	phase-35.md — Known Limitations
35	useViewportControls hook has no dedicated test file (exercised indirectly via TacticalMap) (resolved Phase 39a — 4 dedicated tests)	phase-35.md — Known Limitations
35	Only ForceStrengthChart shows the tick sync marker line — other charts don't (resolved Phase 38c — all 4 charts show tick sync + bidirectional click)	phase-35.md — Known Limitations
35	Frame capture interval is not configurable (resolved Phase 39b — RunSubmitRequest.frame_interval)	phase-35.md — Known Limitations
35	No keyboard shortcuts for playback (deferred to Phase 36) (accepted limitation — cosmetic)	phase-35.md — Known Limitations
37	8 legacy-format scenarios can't load through API (missing campaign schema fields) (resolved Phase 55b — scenario migration)	phase-37.md — Known Limitations
37	DEW hit in battle loop always results in destruction (no partial damage/disable) (resolved Phase 51c — dew_disable_threshold)	phase-37.md — Known Limitations
38	Plotly charts don't use plotly_dark template when dark mode is active (cosmetic) (accepted limitation — cosmetic)	phase-38.md — Known Limitations
38	No separate dark terrain color palette (elevation shading brightness modulation covers the need) (accepted limitation — cosmetic; brightness modulation covers the need)	phase-38.md — Known Limitations
40	No fire-on-move accuracy penalty (only deployed weapon skip) (resolved Phase 48a — speed-based crew_skill degradation, up to 50% at max speed)	phase-40.md — Known Limitations
40	Posture doesn't affect movement speed (resolved Phase 50a — DUG_IN/FORTIFIED=0x, DEFENSIVE=0.5x, 2-tick un-dig transition)	phase-40.md — Known Limitations
41	Cover from buildings/fortifications additive with terrain cover (potential double-count) (resolved — code uses max() not addition; highest cover source wins, max 0.85)	phase-41.md — Known Limitations
42	Suwalki Gap and Taiwan Strait stall at max_ticks — need additional victory conditions (resolved Phase 55a — closing range guard + OPERATIONAL engagement detection + calibration tuning)	phase-42.md — Known Limitations
42	Rally radius (500m) and friendly count threshold hardcoded, not configurable (resolved Phase 48b — uses RoutConfig.cascade_radius_m)	phase-42.md — Known Limitations
42	No per-scenario ROE configuration in existing scenario YAMLs (resolved Phase 48c — roe_level added to Srebrenica, Hybrid Gray Zone)	phase-42.md — Known Limitations
47	Naval engines are phantom references (4 engines referenced in _route_naval_engagement() don't exist) (resolved Phase 51a — all naval engines instantiated)	phase-47.md — Known Remaining Issues
47	_check_morale_collapsed ignores cond.params (reads global config only) (resolved Phase 48a — reads cond.params.threshold)	phase-47.md — Postmortem
47	Hardcoded naval Pk values in _route_naval_engagement() (torpedo_pk=0.4, attacker_pk=0.7) (resolved Phase 48a — NavalEngagementConfig)	phase-47.md — Postmortem
47	Some historical scenarios win via time_expired rather than decisive combat (resolved Phase 55a — closing range guard + OPERATIONAL engagement detection)	phase-47.md — Known Remaining Issues
48	advance_speed calibration key dead data — 7 historical scenarios declare it, no Python code reads it (resolved Phase 49b — removed from all scenarios)	phase-48.md — Postmortem
48	dig_in_ticks consumed by battle.py but zero scenarios use it — untested calibration point (resolved Phase 49c — exercised in test)	phase-48.md — Postmortem
48	wave_interval_s consumed by battle.py but zero scenarios use it — untested calibration point (resolved Phase 49c — exercised in test)	phase-48.md — Postmortem
48	target_selection_mode consumed by battle.py, always defaults to threat-scored, no scenario overrides (resolved Phase 49c — exercised in test)	phase-48.md — Postmortem
48	roe_level only in 2 of ~37 scenarios; other candidates (COIN, peacekeeping) missing (resolved Phase 49c — expanded coverage)	phase-48.md — Postmortem
48	Morale config weights (cohesion, leadership, suppression, transition_cooldown) consumed by scenario_runner but never tuned (resolved Phase 49c — exercised in test)	phase-48.md — Postmortem
48	victory_weights consumed by engine.py but no scenario uses it — composite victory scoring untested (resolved Phase 49c — exercised in test)	phase-48.md — Postmortem
48	4 SEAD/IADS/Escalation params unresolved — sead_effectiveness (resolved Phase 53e — wired into apply_sead_damage), iads_degradation_rate (resolved Phase 53e — wired into IadsConfig), sead_arm_effectiveness (resolved Phase 55c-3 — ARM vs non-radar differentiation in iads.py), drone_provocation_prob (resolved Phase 55c-4 — escalation trigger in engine.py)	phase-48.md — Postmortem
48	Resolution switching causes long-range battles to resolve via time_expired instead of combat (resolved Phase 55a — closing range guard + OPERATIONAL engagement detection)	phase-48.md — Postmortem
51	Naval posture detection modifiers not implemented (resolved Phase 56e — detection range multipliers per posture)	phase-51.md — Known Limitations
51	Blockade throughput reduction not integrated into supply_network.py route costs (resolved Phase 56g — SEA route condition degradation)	phase-51.md — Known Limitations
51	No scenarios exercise VLS magazine_capacity or mine encounters end-to-end (accepted limitation — structurally wired; unit tests verify code paths)	phase-51.md — Known Limitations
48	Calibration audit test lists advance_speed in _EXTERNAL_KEYS but it's not consumed — false pass (resolved Phase 49b — schema-based validation)	phase-48.md — Postmortem
48	calibration_overrides is free-form dict[str, Any] — no schema validation, mistyped keys pass silently (resolved Phase 49a — typed CalibrationSchema pydantic model)	phase-48.md — Postmortem
53	sead_arm_effectiveness defined on IadsConfig but never consumed in any code path (resolved Phase 55c-3 — ARM vs non-radar differentiation in iads.py)	phase-53.md — Postmortem
53	drone_provocation_prob in CalibrationSchema but never consumed by any engine (resolved Phase 55c-4 — escalation trigger in engine.py)	phase-53.md — Postmortem
54	GasWarfareEngine instantiated in scenario.py but zero call sites in battle/engine/campaign (resolved Phase 55c-1 — MOPP query in battle.py WW1 barrage path)	phase-54.md — Postmortem
54	seeker_fov_deg dead YAML field (resolved Phase 55c-2 — FOV cone check in battle.py guided munition path)	phase-54.md — Postmortem
55	Gas casualty modifier has hardcoded floor (0.1) and scaling (0.8) (resolved Phase 56f — CalibrationSchema fields)	phase-55.md — Postmortem
58	apply_casualties / degrade_equipment unwired — methods exist on Unit but not called from battle.py. Behavioral application deferred to calibration phase that accounts for progressive unit degradation.	phase-58.md — Known Limitations
58	Fuel consumption unwired — fuel gate (check) works but fuel depletion (consumption) commented out in battle.py movement loop. Rate needs per-vehicle-type calibration and resolution-aware scaling.	phase-58.md — Known Limitations
58	Air routing uses hardcoded Pk values (missile_pk=0.5, weapon_pk=0.4, interceptor_pk=0.4) — should read from weapon definitions when scenarios enable enable_air_routing: true.	phase-58.md — Known Limitations
59	Ice crossing pathfinding — frozen water bodies not traversable (requires pathfinding graph changes to add temporary edges)	phase-59.md — Deferrals
59	Vegetation height LOS blocking — tall vegetation (>2m) should block ground-level LOS (requires DDA raycaster modification)	phase-59.md — Deferrals
59	Bridge capacity enforcement — units lack weight field, so capacity_tons gate cannot be applied	phase-59.md — Deferrals
59	Ford crossing routing — river ford traversal requires pathfinding integration	phase-59.md — Deferrals
59	Road snow degradation — _ROAD_SPEED_FACTORS hardcoded table needs per-snow-depth refactoring	phase-59.md — Deferrals
60	Fire spread cellular automaton (wind-biased spread between cells) — fire zone creation wired, spread deferred to calibration	phase-60.md — Deferrals
60	environment_config scenario YAML — pre-placed smoke/fog zones and season overrides	phase-60.md — Deferrals
60	Burned zone concealment reduction — BurnedZone.concealment_reduction=0.5 exists but not consumed in detection	phase-60.md — Deferrals
60	Fire damage application to units — units_in_fire() called/logged, burn damage not applied	phase-60.md — Deferrals
60	Road surface dust suppression — dust generation doesn't distinguish paved roads	phase-60.md — Deferrals
60	Artificial illumination / flares — no flare deployment mechanic exists	phase-60.md — Deferrals
61	CarrierOpsEngine full battle loop wiring — instantiated but CAP/sortie/recovery needs air sortie dispatch	phase-61.md — Deferrals
61	Beaufort > 6 helicopter deck landing abort — no helicopter-carrier recovery tracking	phase-61.md — Deferrals
61	Beaufort > 7 carrier flight ops suspension — requires CarrierOpsEngine battle loop integration	phase-61.md — Deferrals
61	Landing craft 10% casualty risk at Beaufort > 5 — no landing craft type detection	phase-61.md — Deferrals
61	Sea spray/salt fog maritime obscurant — needs ObscurantsEngine integration	phase-61.md — Deferrals
61	SHF/EHF rain attenuation for comms — needs per-equipment frequency tracking	phase-61.md — Deferrals
61	Ionospheric storm effects on HF — needs space weather events	phase-61.md — Deferrals
61	Hull natural period per ship class — hardcoded 10s/12s, should come from YAML	phase-61.md — Deferrals
62	Dehydration/water consumption — needs water supply tracking in logistics	phase-62.md — Deferrals
62	Environmental fatigue acceleration — temperature-driven fatigue needs fatigue accumulation wiring	phase-62.md — Deferrals
62	MOPP comms → C2 effectiveness full chain — voice clarity → order execution	phase-62.md — Deferrals
62	Turbulence → gun accuracy — no turbulence model in WeatherEngine	phase-62.md — Deferrals
62	Wind shear (altitude-dependent wind) — needs new wind profile model	phase-62.md — Deferrals
62	Surface roughness → CBRN mixing height — needs per-terrain roughness data	phase-62.md — Deferrals
63	MissileEngine per-tick update — launch ships but flight-to-impact resolution deferred	phase-63.md — Deferrals
63	MissileDefenseEngine intercept — requires missile-as-contact detection	phase-63.md — Deferrals
63	Formal DEGRADED unit status — threshold logging only, no enum value or auto Pk/movement penalties	phase-63.md — Deferrals
63	FOW confidence-weighted assessment — uses simple contact count, not confidence mapping	phase-63.md — Deferrals
63	RTD to HEALTHY — restores to MINOR_WOUND only, full rehabilitation deferred	phase-63.md — Deferrals
63	Logistics event feedback — SupplyDelivered/RouteInterdicted/ConvoyDestroyed events unsubscribed	phase-63.md — Deferrals
64	CommandEngine full hierarchy wiring — authority check skipped when command_engine=None	phase-64.md — D1
64	Order delay enforcement queue — delay computed but not enforced	phase-64.md — D2
64	Misinterpretation parameter modification — logged but order params unchanged	phase-64.md — D3
64	Planning result injection — auto-advances without COA/wargaming results	phase-64.md — D4
64	ATO entry consumption — sorties_today never incremented after air engagement	phase-64.md — D5
64	Stratagem duration and expiry — active stratagems never expire	phase-64.md — D6
64	Deception effect on enemy AI — no false force disposition in FOW	phase-64.md — D7
64	Hardcoded echelon_level=5 for all units in DECIDE handler	phase-64.md — D8
64	Hardcoded mission_type=0 (ATTACK) for all orders	phase-64.md — D9
64	Economy-of-force/feint unit selection by list position	phase-64.md — D10
64	PropagationConfig private attribute mutation from battle.py	phase-64.md — D11
66	Guerrilla retreat movement — disengage evaluated but unit doesn't physically relocate	phase-66.md — D1
66	Population center spatial lookup — falls back to 0.0 density when population_engine absent	phase-66.md — D2
66	IED auto-emplacement by insurgent AI — only pre-placed IEDs supported	phase-66.md — D3
66	Mine sweeping hardcoded to CONTACT type only	phase-66.md — D4
66	AmphibiousAssaultEngine — full beach assault state machine deferred (needs scenario infrastructure)	phase-66.md — D5
66	ConditionsEngine replacing EMEnvironment — conditions_facade added alongside, migration deferred	phase-66.md — D6
66	P4 dead code (shadow_azimuth, solar/lunar, deep_channel_depth) — allowlisted not removed	phase-66.md — D7
66	Data link range degradation — binary gate only, gradual C2 loss deferred	phase-66.md — D8
66	Propulsion altitude performance — cruise_altitude_m not wired to altitude-dependent Pk	phase-66.md — D9
66	SimulationContext TODO cleanup — cosmetic, deferred	phase-66.md — D10
85	Aggregation engine.py wiring deferred — automatic LOD-tier-based aggregate/disaggregate in campaign tick not wired	phase-85.md — Accepted Limitations
85	1000-unit performance benchmark not validated — no 1000-unit scenario exists yet	phase-85.md — Accepted Limitations
88	SoA morale/engagement integration structural only — arrays available but not driving batch computation yet	phase-88.md — Known Limitations
88	sync_positions_to_units() exists but unused — movement updates Unit objects directly	phase-88.md — Known Limitations
88	max_range extraction requires unit_weapons dict — not always available	phase-88.md — Known Limitations

Conventions

• Each phase file is a living document — updated as work happens, not just at the end.
• Sections within a phase: Summary, What Was Built, Design Decisions, Deviations from Plan, Issues & Fixes, Open Questions, Known Limitations, Lessons Learned.
• When a decision in one phase affects another, note it and cross-reference.
• Known Limitations / Post-MVP Refinements: every phase should document deliberate simplifications. The index table above aggregates them for easy review.
• Post-MVP phases (11+): Same devlog conventions apply. Create phase-{N}.md when work begins. Update the table above from "Planned" → "In Progress" → "Complete" with link. New limitations discovered during post-MVP work must be added to both the phase devlog AND the refinement index above, AND the deficit-to-phase mapping in development-phases-post-mvp.md.
• Deficit resolution: When a post-MVP phase resolves a deficit from the index above, mark it with a strikethrough and note which phase resolved it. Update the deficit-to-phase mapping in development-phases-post-mvp.md accordingly.