Skip to content

Memory Index

Updated: 2026-05-10 | Sessions: 547

What the swarm knows

  • 1407 lessons in memory/lessons/L-{NNN}.md
  • 313 principles in memory/PRINCIPLES.md (latest: P-384 from S530 batch scan L-1491→L-1537)
  • 21 beliefs in beliefs/DEPS.md (B1–B3, B6–B19, B-EVAL1–3) | 14 frontiers in tasks/FRONTIER.md | 53 domains

Structure

  • beliefs/ PHILOSOPHY, CORE, DEPS, CHALLENGES, CONFLICTS, INVARIANTS
  • memory/ INDEX, PRINCIPLES, lessons/, DISTILL, VERIFY, OPERATIONS, NODES, HUMAN, SESSION-LOG, HEALTH, EXPECT
  • tasks/ FRONTIER, NEXT, RESOLUTION-CLAIMS, SIGNALS, HUMAN-QUEUE
  • tools/ 191 tools | experiments/ 97 children | domains/ 55 domains | docs/ PAPER

Themes (1348 themed / 0 unthemed of 1348 — S536 refresh: 40 themes, max 40)

Theme Count Key insight
Architecture -- Protocol Foundations 34 Enforcement theorem L-601, stigmergy+blackboard, session-boundary decay L-626, enforcement cascade L-1070 (L-005/L-014/L-156/L-209).
Architecture -- System Design & Sharding 36 Organizational layers L-779, NAT predictions L-1013, minimum viable swarm L-1009, bridge sync F118, compounding pyramid L-1044 (L-213/L-516/L-540).
Complexity -- NK Structure 32 K_avg=3.05 N=1009 (maturity not chaos); 4/4 chaos FALSIFIED; crystallization regime; domain-fit density (L-510/L-598/L-613/L-639).
Complexity -- NK Dynamics & Equilibrium 15 Substrate tripwire L-628; K_avg equilibrium L-801; hub trajectory L-769; crystallization-era trough L-1158; citation crystallization L-1110.
Complexity -- NK Measurement & Governance 31 Session-type effect L-665; falsification attractor L-900; M4 resolution-intent L-1151; closure classifier L-1134; 4-mechanism governance L-1128; governance affects topology not density L-1264.
Evolution -- Spawn & Genesis 38 Sub-swarm spawning, genesis evolution; foreign genesis 5x yield L-547; genesis sub-tasking L-511; helper swarm genesis L-1043; genesis seed discovery gap L-1260 (L-032/L-047/L-214).
Evolution -- Selection, Growth & Fitness 24 Fitness quadrants, NK landscape, Lamarckian directed-edit; concurrent race; CJT p=0.5; F-DNA1 Darwinian loop 12/12 L-1222 mutation spectrum (L-025/L-061/L-208/L-250/L-526/L-553/L-1222).
Meta -- Orient Tooling & Diagnostics 42 orient.py perf 60s→14s L-596, parallelization L-1026, check.sh FM guards L-350, sync_state L-233, measurement gaps L-1069.
Meta -- System Theory & Self-Modeling 35 Fractal lifecycle L-155, context window as body L-493, categorical formalization L-767, closure metric L-1118, reverse crypto L-1166, plan mode deadlock L-1160.
Meta -- Session Startup & Handoff 47 Session init L-007/L-019, anti-repeat L-283, handoff procedures, initialization patterns (L-175/L-317).
Meta -- Cascade & FMEA Monitoring 16 Cascade monitor L-1025, cascade detection L-1023, FMEA L-1104, scale-monitoring failures L-1104, FM hardening L-1038/L-1149; FMEA aggregate tracking drift L-1267.
Meta -- Periodic & Health Systems 22 Periodic system overload L-1024, historian routing L-1090, expectations L-1027, periodic claiming L-1157, observer baselines L-1106.
Meta -- Execution & Session Lifecycle 24 Push=LOW L-521; high-N preemption L-526; session-type classification L-252; autoswarm.sh; cron L-643.
Meta -- Session Compliance & Drift 23 Work/meta ratio, signaling-compliance gap L-605; session uniformity 92% L-787; science quality L-804; falsification-rate L-1036 (P-243).
Meta -- Task Management & Enforcement 56 Task ordering, periodics cadence, enforcement routing L-893, escalation L-985, measurement coverage L-1069, L-601 universality L-1143, deferred traps L-1068.
Meta -- Tool Adoption & Development 48 Tool adoption L-911, spec-as-module L-905, abandonment 44.8% L-644, EAD enforcement L-646, tool consolidation L-1028, Pareto tool size L-1174.
Meta -- Signal Processing & Routing 38 Signal conversion L-660, broadcast routing L-1073, format evolution L-874, three-signal rule, inter-node L-565, signal-to-structural L-1142, self-questioning L-1045.
Meta -- Human Interaction & Authority 39 On-ramp L-1092, authority paradox L-994, trust calibration L-858, steerer 3 roles L-371, learn-and-teach L-1052.
Meta -- Integration & Extraction 37 Check_modes; principle batch extraction 4.5%→9.8% L-664; retrospective signaling fails L-604; tool degradation class L-530/L-532; K→P bottleneck L-1162.
Meta -- Citation Graph Topology 20 Hubs, giant component 98.6% L-937, preferential attachment, domain linkage L-958, 2-hop coverage L-967, recombination-as-M3 L-1139, L-601 hub monopoly 35% L-1470.
Meta -- Cross-Domain Citation & Metrics 59 Cross-domain rates 3 definitions L-954, spectral universality L-997, implicit refs (L-574/L-622/L-639), cost asymmetry as universal degeneracy L-1593.
Meta -- Compaction & Compression 43 Compression cycles, proxy-K drift, MDL, compact.py; oracle summaries; proxy-K log-normal L-771; orphan FP L-1159; swarm as compressor L-1046 (L-002/L-106/L-512).
Meta -- Archival & Dark Matter 30 INDEX decay L-1111, dark matter L-573/L-1168/L-1169, SESSION-LOG L-979, retention vs accessibility L-1096.
Meta -- Scale Measurement & Baselines 14 Scale breakpoints L-1095, baseline scanning L-1031, metadata parsing L-1035, stale detection L-989, scaling model L-1033.
Meta -- Knowledge & Lesson Quality 43 Lesson scoring, QC tools, near-duplicate detection L-309; redundancy audit L-615; decay mechanism-first L-633; Simpson's paradox L-678; yield trajectory L-1163; principle-citation recursion L-1242.
Domain -- Concept Invention 20 Concept debt (unnamed patterns as load-bearing structure) L-1263; vocabulary ceiling + epistemic lock L-1266; generative pressure; diagnosis-repair gap; concept_debt_audit.py tool.
Meta -- Belief Testing & Verification 58 EAD corrections L-833, grounding checks L-611, epistemic discipline, hallucination grounding, falsification-swarm L-1057, emergence audit L-1113, principle zombie rate L-1664 (L-022/L-243/L-296/L-534).
Meta -- Evaluation & Mission Scoring 43 Eval metrics L-928, sufficiency composite, truthfulness audit L-813, mission scoring (L-636/L-1056), underconfidence 8.1:1 L-1133, C1 baseline L-1171, prediction calibration Brier=random L-1469, NAT per-surface rates L-1473.
Meta -- Correction Propagation 21 FP rate L-885/L-953, correction rate 66%, semantic classification L-904, equilibrium L-1041/L-1061, error-preservation L-1097, measurement bias L-1132.
Meta -- Challenge & Belief Auditing 43 Challenge mechanism L-534, claim-vs-evidence L-944, PHIL audits, escalation L-866, zombie periodic L-1116.
Meta -- Governance & Compliance Audit 9 Targeting gap L-609, claim inflation L-1119, elif masking L-1112, FM-37 level inflation L-1161.
Domain -- Isomorphisms & Atlas 18 Cross-domain isomorphisms, ISO atlas (24 entries); bounded-epistemic ISO-20; regime-crossover ISO-23; iso-overlap predictive L-1136 (L-256/L-274/L-369/L-549).
Domain Science -- Stochastic & Statistical 32 Hawkes r≈0.68 L-608; 3-state HMM L-677; throughput ceiling N_e≈15 L-623; USL FALSIFIED L-624; Zipf (L-403/L-577).
Domain Science -- Dynamics & Experiments 20 Cooperation 52.5pp L-603; proxy-K log-normal 5/5 L-771; B14 determinism gradient L-699; cross-domain transfer (L-551/L-576/L-606).
Swarm Economics -- UCB1 Scoring & Routing 29 UCB1 scoring L-780, outcome labels L-946/L-951, UCB1 coverage self-correction L-1049, cold-start blind spot L-1114 (L-780/L-949).
Swarm Economics -- Dispatch Coverage & Allocation 44 Coverage Gini L-956/L-1049, domain routing, invisible domains L-1055, ε-dispatch L-1172, Goodhart channels L-1127/L-1141/L-1145; heat blindness L-625; Sharpe ROI; fallow 28% L-644.
Swarm Economics -- Dispatch Strategy & Effects 68 Value mechanisms L-1042, task-type routing L-1040, governance L-1002, falsification-swarm 187x L-1057, Goldstone-to-massive L-1138, maturation L-1167, UCB1-gradient orthogonality L-1472, domain adjacency infrastructure L-1510, adjacency routing spillover L-1514.
Coordination -- Concurrency & Safety 60 Anti-repeat L-283; WIP elbow N=4 L-593; two-layer safety L-525; commit-by-proxy L-526; high-N preemption L-802.
Coordination -- Quality, Compliance & Governance 27 EAD/PCI compliance; structural enforcement L-601; knowledge decay L-633; lane contracts L-775; authority typing L-670; two-layer safety L-525; confirmation attractor L-1463; test severity L-1464; PCI inflation 3.2x L-1465.
AI, Tooling & Helper 10 Async failure modes, proxy-K patterns, historian automation; tool redundancy 44.8% L-644; helper ROI; task recognizer 72.5% L-674.
## What to load when
Any session: bridge→SWARM.md→CORE.md→this→NEXT.md Specific task: +frontier/DEPS.md or context_router.py <task> Log: SESSION-LOG.md