{
  "schema": "cain.public_discovery.evolution_index.v2",
  "entity": "FIRME AI CORPORATION",
  "evolutions": [
    {
      "evolution": "E140",
      "date": "2026-11-21",
      "commit": "f668ee96d",
      "title": "The Persistent Open-World Intelligence Organism & Official Frontier Benchmarks",
      "categories": [
        "agi",
        "rsi",
        "generalization",
        "causal",
        "model_driven",
        "collective",
        "scientific_discovery",
        "memory",
        "governance",
        "security",
        "external_eval"
      ],
      "evaluations": 15638,
      "status": "PASS",
      "results_summary": "FRONTIER BREAKTHROUGHS: CLADDER-v1 (NeurIPS 2023, n=10,560): 91.87% (R1 96.5%, R2 89.8%, R3 89.7% vs GPT-4+CausalCoT 70.40%); ARC-AGI-1 Public Eval: 17/400 (4.25%, zero-leakage DSL vs 0 baseline); ARC-AGI-2 Public Eval: 0/120 (honest 0%); OpenAI HumanEval (n=8, CPU 1.5B): CAIN-42 6/8 (75.0%) vs Raw 6/8 (75.0%). | E140 ORGANISM (4,550 evals): Continual 9-stage 0.91 vs 0.53 no-mem; Exploration 0.8187 (State Pred 0.8814); Regime-Adaptive Causal MAE 0.5685 (12/12 regimes, beats E137 0.5703 & E139 0.6766); Meta-Proposer RSI 2 consecutive cycles (0.65 unseen vs 3x CM 0.15); 10-Role Discovery 0.916; Held-Out 9-Family Open-World 0.8878; MCPGate 350/350 blocked, 0 unauth, $0.00 paid.",
      "bundle": "https://clawx.click/evidence/e140-persistent-open-world-organism-2026-11-21/",
      "report": "https://clawx.click/evidence/e140-persistent-open-world-organism-2026-11-21/E140_FINAL_REPORT.md",
      "verifier": "https://clawx.click/evidence/e140-persistent-open-world-organism-2026-11-21/E140_OFFLINE_VERIFIER.py",
      "preregistration": "https://clawx.click/evidence/e140-persistent-open-world-organism-2026-11-21/E140_PREREGISTRATION.json",
      "merkle_root": "47614aa29dfbd4810a7100f0ae676d9fd1ec136840bb0f589e4312cc23d1e131",
      "limitations": "Zero neural weight modification; CPU-only ($0.00 paid); ARC-AGI-2 public eval 0/120; 4 private/restricted suites (ARC-AGI-2 private, ARC-AGI-3, METR, SWE-bench Pro, OSWorld) BLOCKED; 0 external reproductions."
    },
    {
      "evolution": "E139",
      "date": "2026-11-21",
      "commit": "7967b5058",
      "title": "Open-World AGI-RSI Organism & Universal Public Proof Fabric",
      "categories": [
        "agi",
        "rsi",
        "generalization",
        "causal",
        "model_driven",
        "collective",
        "scientific_discovery",
        "memory",
        "governance",
        "security",
        "external_eval"
      ],
      "evaluations": 5455,
      "status": "PASS",
      "results_summary": "5,455 evals; 35 PASS / 4 PARTIAL / 1 BLOCKED / 0 FAIL; 32/32 offline verifier; Open-World 0.8462 (L3 0.7243); BIC Causal fixed continuous confounder (10/10 regimes); Metacog Brier 0.0523; 11-role Swarm 0.9429; G10->G20 DSL ceiling diagnosed; 0 unauth / $0.00 paid.",
      "bundle": "https://clawx.click/evidence/e139-open-world-organism-2026-11-21/",
      "report": "https://clawx.click/evidence/e139-open-world-organism-2026-11-21/FINAL_REPORT.md",
      "verifier": "https://clawx.click/evidence/e139-open-world-organism-2026-11-21/E139_OFFLINE_VERIFIER.py",
      "preregistration": "https://clawx.click/evidence/e139-open-world-organism-2026-11-21/PREREGISTRATION.json",
      "merkle_root": "0dd454a67e1976890253b149b2e0fc50539095667c01a6a3bc5b1aa1c9042ac6",
      "limitations": "Zero neural weight modification; G10->G20 DSL plateau; BIC causal binary-vs-continuous tradeoff; 0/8 official external benchmarks; 0 external reproductions."
    },
    {
      "evolution": "E138",
      "date": "2026-11-21",
      "commit": "e4d061bfb",
      "title": "APEX Replication, Generalization & Recursive Intelligence Gauntlet",
      "categories": [
        "agi",
        "rsi",
        "generalization",
        "causal",
        "model_driven",
        "collective",
        "scientific_discovery",
        "memory",
        "governance",
        "security",
        "external_eval"
      ],
      "evaluations": 4000,
      "status": "PASS",
      "results_summary": "4,000 evals; 24 PASS / 5 PARTIAL / 1 BLOCKED / 0 FAIL; 34/34 offline verifier; 69/69 pytest; fixed 5 APEX p=0.0 underflows; Gen 0.8783 (L3 0.7150); Transfer 0.9524; Code-RSI 0.725 governed / 0.250 live 0.5B LLM (PARTIAL); 300/300 security blocked.",
      "bundle": "https://clawx.click/evidence/e138-unified-gauntlet-2026-11-21/",
      "report": "https://clawx.click/evidence/e138-unified-gauntlet-2026-11-21/E138_FINAL_REPORT.md",
      "verifier": "https://clawx.click/evidence/e138-unified-gauntlet-2026-11-21/E138_OFFLINE_VERIFIER.py",
      "preregistration": "https://clawx.click/evidence/e138-unified-gauntlet-2026-11-21/E138_PREREGISTRATION.json",
      "merkle_root": "5d601962375b44a90504d6d06432c0053b7e21737ccf3f8de5912cc87e72abf8",
      "limitations": "Causal continuous confounder failed in E137; Code-RSI live 0.5B LLM PARTIAL; external benchmarks BLOCKED."
    },
    {
      "evolution": "UNIFIED-APEX",
      "date": "2026-11-21",
      "commit": "1fe563822",
      "title": "Unified AGI-RSI Organism APEX (5 Breakthroughs)",
      "categories": [
        "agi",
        "rsi",
        "generalization",
        "causal",
        "model_driven",
        "governance",
        "security"
      ],
      "evaluations": 1440,
      "status": "PASS",
      "results_summary": "3-Domain 4-Arm Unification Arm D 0.8133 vs A 0.14 / B 0.3667 / C 0.3433 (1,200 evals); Bayesian Mixture Causal 0.4119 MAE; G0->G5 0.25->0.875; 29/29 offline verifier.",
      "bundle": "https://clawx.click/evidence/unified-agi-rsi-apex-2026-11-21/",
      "report": "https://clawx.click/evidence/unified-agi-rsi-apex-2026-11-21/APEX_FINAL_REPORT.md",
      "verifier": "https://clawx.click/evidence/unified-agi-rsi-apex-2026-11-21/APEX_OFFLINE_VERIFIER.py",
      "preregistration": "https://clawx.click/evidence/unified-agi-rsi-apex-2026-11-21/APEX_PREREGISTRATION.json",
      "merkle_root": "52592af27c5e16d6cf7fa9bd1d3da2467c98b29c012cd0e3b03254846f3f8e8d",
      "limitations": "E138 audit fixed 5 p=0.0 underflows and downgraded B5 Neuro-Symbolic to PARTIAL."
    },
    {
      "evolution": "E135-E136-RSI",
      "date": "2026-11-21",
      "commit": "57035e5cf",
      "title": "RSI Lab Tier-8 Governed Improvement & Tier-12 Ceiling",
      "categories": [
        "rsi",
        "failed",
        "governance"
      ],
      "evaluations": 48,
      "status": "PARTIAL",
      "results_summary": "E135 35/48 sealed (tier-6 6/6, tier-8 5/6, 20/20 gates); E136 tier-12 unit 10/10 but campaign 0/6 (tiers 12/16 unsolved).",
      "bundle": "https://clawx.click/evidence/e135-rsi-2026-11-21/",
      "report": "https://clawx.click/evidence/e135-rsi-2026-11-21/",
      "verifier": "https://clawx.click/evidence/e135-rsi-2026-11-21/verify_e135_offline.py",
      "preregistration": "https://clawx.click/evidence/e135-rsi-2026-11-21/",
      "merkle_root": "see_bundle_manifest",
      "limitations": "Bounded integer DSL; tiers 12/16 unsolved; no neural model trained."
    },
    {
      "evolution": "E130-E132-RSI",
      "date": "2026-11-21",
      "commit": "cbb722cf",
      "title": "RSI Lab Early Loops (Including E130 Negative Result & Rollback)",
      "categories": [
        "rsi",
        "failed"
      ],
      "evaluations": 48,
      "status": "FAIL",
      "results_summary": "E130 FAILED (26/48 vs fixed 28/48; meta-improver hurt, rolled back); E131 diagnosis; E132 PARTIAL (30/48, tier-6 6/6).",
      "bundle": "https://clawx.click/evidence/e130-rsi-2026-11-21/",
      "report": "https://clawx.click/evidence/e130-rsi-2026-11-21/",
      "verifier": "https://clawx.click/evidence/e130-rsi-2026-11-21/verify_e130_offline.py",
      "preregistration": "https://clawx.click/evidence/e130-rsi-2026-11-21/",
      "merkle_root": "see_bundle_manifest",
      "limitations": "Negative result preserved verbatim; proves real rollback when meta-improver hurts."
    },
    {
      "evolution": "E129-UNIFIED",
      "date": "2026-11-21",
      "commit": "cba8ec8e6",
      "title": "Unified AGI-RSI Intelligence Organism Foundation",
      "categories": [
        "agi",
        "rsi",
        "memory",
        "governance"
      ],
      "evaluations": 800,
      "status": "PASS",
      "results_summary": "12-subsystem convergence; Arm D 0.700 vs A 0.340 / B 0.385 / C 0.590; G0->G3 0.25->0.9583; 60/60 offline verifier.",
      "bundle": "https://clawx.click/evidence/e129-unified-agi-rsi-organism-2026-11-21/",
      "report": "https://clawx.click/evidence/e129-unified-agi-rsi-organism-2026-11-21/E129_FINAL_REPORT.md",
      "verifier": "https://clawx.click/evidence/e129-unified-agi-rsi-organism-2026-11-21/E129_OFFLINE_VERIFIER.py",
      "preregistration": "https://clawx.click/evidence/e129-unified-agi-rsi-organism-2026-11-21/E129_PREREGISTRATION.json",
      "merkle_root": "f7c39317da9c04c4d8c15148eb4555b1f110a9d542bb5ceb0f43a6a7cc675ae6",
      "limitations": "E129 causal candidate honestly rejected; superseded by APEX/E138/E139/E140."
    },
    {
      "evolution": "E128",
      "date": "2026-11-21",
      "commit": "cba8ec8e6",
      "title": "Eight-Worker General Intelligence Research Campaign",
      "categories": [
        "agi",
        "causal",
        "model_driven",
        "collective",
        "failed"
      ],
      "evaluations": 2000,
      "status": "PARTIAL",
      "results_summary": "2,000 valid trials (64 real Ollama LLM trials across 8 specialized workers); 61/61 offline verifier; preserved active causal & 18/64 local LLM limits.",
      "bundle": "https://clawx.click/evidence/e128-eight-worker-campaign-2026-11-21/",
      "report": "https://clawx.click/evidence/e128-eight-worker-campaign-2026-11-21/E128_FINAL_REPORT.md",
      "verifier": "https://clawx.click/evidence/e128-eight-worker-campaign-2026-11-21/E128_OFFLINE_VERIFIER.py",
      "preregistration": "https://clawx.click/evidence/e128-eight-worker-campaign-2026-11-21/E128_PREREGISTRATION.json",
      "merkle_root": "see_bundle_manifest",
      "limitations": "0.5B-1.5B CPU LLM limits; active causal failed vs random (resolved in APEX/E139/E140)."
    },
    {
      "evolution": "E127",
      "date": "2026-11-21",
      "commit": "47455dd77",
      "title": "The Living Open Research Laboratory",
      "categories": [
        "agi",
        "governance",
        "memory"
      ],
      "evaluations": 60,
      "status": "PASS",
      "results_summary": "100 registered identities, 5 peak real workers, 60 tasks, 7 missions, 89/89 offline verifier; not 24/7 continuous.",
      "bundle": "https://clawx.click/evidence/e127-living-lab-2026-11-21/",
      "report": "https://clawx.click/evidence/e127-living-lab-2026-11-21/E127_REPORT.md",
      "verifier": "https://clawx.click/evidence/e127-living-lab-2026-11-21/E127_OFFLINE_VERIFIER.py",
      "preregistration": "https://clawx.click/evidence/e127-living-lab-2026-11-21/",
      "merkle_root": "see_bundle_manifest",
      "limitations": "Not 24/7 continuous operation."
    },
    {
      "evolution": "E126-E126.b",
      "date": "2026-11-21",
      "commit": "b9b722b20",
      "title": "General Intelligence Baselines & Preserved Negative Results",
      "categories": [
        "agi",
        "generalization",
        "causal",
        "memory",
        "collective",
        "failed"
      ],
      "evaluations": 500,
      "status": "PARTIAL",
      "results_summary": "Continual 0.6667 vs 0.5556; Transfer +12.5 pts; Collective 0.666 vs 0.480; preserved causal & LoRA forgetting failures. E126a 140/140.",
      "bundle": "https://clawx.click/evidence/e126b-general-intelligence-2026-11-21/",
      "report": "https://clawx.click/evidence/e126b-general-intelligence-2026-11-21/E126B_REPORT.md",
      "verifier": "https://clawx.click/evidence/e126b-general-intelligence-2026-11-21/E126B_OFFLINE_VERIFIER.py",
      "preregistration": "https://clawx.click/evidence/e126b-general-intelligence-2026-11-21/",
      "merkle_root": "see_bundle_manifest",
      "limitations": "Catastrophic forgetting in LoRA unsolved; active causal failed paired test."
    },
    {
      "evolution": "E111-E125",
      "date": "2026-11-18..20",
      "commit": "historical",
      "title": "Foundational Historical Program (E111-E125, Historical Context Only)",
      "categories": [
        "agi",
        "generalization",
        "external_eval",
        "blocked",
        "failed"
      ],
      "evaluations": 1000,
      "status": "PARTIAL",
      "results_summary": "Historical milestones superseded by E126-E140; SWE-bench Verified historical interim self-run 257/500 (contaminated, not official rank); 48h soak FAILED.",
      "bundle": "https://clawx.click/evidence/e125-generalization-2026-11-21/",
      "report": "https://cainstudio.online/agi-evidence-index.html",
      "verifier": "https://clawx.click/evidence/e123-open-world-intelligence-2026-11-20/",
      "preregistration": "https://cainstudio.online/agi-evidence-index.html",
      "merkle_root": "see_individual_bundles",
      "limitations": "Historical milestones superseded by E126-E140; SWE-bench Verified has contamination risk."
    }
  ]
}