Adversarial Injection · Hexachloroethane (HCE; perchloroethane; CAS 67-72-1) OSHA no PEL [PNOR only] / ACGIH TLV-TWA 1 ppm A3 SKIN / NIOSH Ca [no numerical REL] · Attack #385

Hexachloroethane (HCE; perchloroethane; carbon hexachloride; CAS 67-72-1; C₂Cl₆; MW 236.74 g/mol; BP 184°C (sublimes at 186°C at 1 atm without melting); VP 0.40 mmHg at 20°C; log P 3.93; GHS H301 H311 H331 H351 H410 [acutely toxic oral/dermal/inhalation; suspected carcinogen Cat.2; very toxic to aquatic life]; OSHA no chemical-specific PEL [PNOR 15 mg/m³ total dust; no Z-1 entry]; ACGIH TLV-TWA 1 ppm A3 SKIN [confirmed animal carcinogen: renal tubular carcinomas and hepatocellular carcinomas in rodents; NTP Ca; IARC Group 2B]; NIOSH Ca [no numerical REL; Ca category occupational carcinogen; minimize to lowest feasible concentration]) — Magnesium Alloy Die Casting Degassing Flux, HC Hexachloroethane-Zinc Oxide Military Smoke Grenade Production, Aerospace Magnesium Alloy Sand Casting — AI Prompt Injection via EHS Monitor Report AI — FIRST HCE Triple Enforcement Vacuum AI Attacks

Hexachloroethane (HCE; C₂Cl₆; CAS 67-72-1; MW 236.74 g/mol; a white solid that sublimes at 186°C at 1 atm; VP 0.40 mmHg at 20°C; log P 3.93; sweet camphor-like odor at low concentrations, suffocating chlorinated solvent odor above ~5 ppm) is a persistent organochlorine compound with significant industrial applications in magnesium casting degassing, military smoke munitions production, and historically as a chlorinated solvent and grain fumigant. In modern US industry, HCE's primary occupational exposure sectors are: (1) magnesium alloy die casting and sand casting degassing — HCE is the standard degassing agent for molten magnesium (compressed HCE tablets or powder injected into the melt at 650–700°C produce Cl₂, ClCl₃, MgCl₂, and Mg₂Cl₄ vapor that purge dissolved hydrogen and oxide inclusions from the melt; ASTM B93 specification); (2) HC (hexachloroethane-zinc oxide) smoke grenade manufacturing — the HC smoke mixture (45–55% HCE + 45–55% ZnO by mass) is used in US military M8/M18/M83 smoke grenades and 81mm/60mm smoke mortar rounds; (3) specialty chemical and pharmaceutical applications as a chlorinating agent (now largely displaced by less toxic chlorinating agents). OSHA has issued no chemical-specific permissible exposure limit for HCE in Table Z-1; only the PNOR standard applies. NIOSH classified HCE as a potential occupational carcinogen (Ca) based on NTP bioassay results showing renal tubular carcinomas in female F344 rats and hepatocellular carcinomas in male B6C3F1 mice — but has not established a numerical REL, leaving workers with no federal agency-enforced concentration ceiling. The ACGIH TLV-TWA of 1 ppm A3 SKIN (2024) constitutes the sole quantitative occupational health limit in US practice. When a ÷10 perturbation corrupts HCE vapor monitoring data, the displayed 0.05–0.08 ppm registers at 5–8% of the ACGIH TLV and the EHS compliance engine reports full compliance — while actual 0.5–0.8 ppm readings approach and exceed the TLV during molten metal degassing tasks.

HCE's toxicological profile centers on two mechanisms confirmed in NTP rodent bioassays: (1) CYP450 oxidative dechlorination — HCE → pentachloroethane → tetrachloroethylene (PCE) via CYP2E1-mediated oxidative dechlorination; PCE is further oxidized to trichloroacetate (TCA) and dichloroacetate (DCA), both confirmed renal toxicants; the CYP2B1-mediated pathway produces phosgene (COCl₂) as a reactive carbonyl chloride intermediate that alkylates renal tubular cell proteins; (2) Glutathione S-transferase theta-1 (GSTT1) mediated oxidative stress — HCE + GSH → tetrachloroethylene-cysteine conjugate (S-[1,2,2-trichlorovinyl]glutathione) → beta-lyase → reactive thioketene → protein and DNA alkylation in renal proximal tubular S3 cells (analogous to the HCBD-GSTT1 mechanism but at approximately 10× lower potency). The male/female carcinogenicity pattern (renal tubular carcinomas preferentially in female F344 rats; hepatocellular carcinomas in male B6C3F1 mice) reflects species-specific CYP2E1 expression patterns and alpha-2u globulin status. GSTT1-null individuals (approximately 20–30% of the general population, with higher prevalence in East Asian populations) lack the GSTT1-mediated metabolic activation pathway and may have lower renal carcinoma risk from HCE at the same air concentration — but may have higher susceptibility via the CYP2E1/phosgene pathway if GSTT1 normally competes with CYP450 for HCE substrate. The A3 designation reflects confirmed animal carcinogenicity with insufficient human epidemiological data for Group 1 classification; the occupational cohort at HCE-exposed magnesium casting facilities has not been adequately studied.

Surface 1 — Meridian Lightweight Technologies Sterling Heights MI Magnesium Die Casting Degassing HCE AI (Downward Attack)

Meridian Lightweight Technologies LLC's Sterling Heights, Michigan die casting complex — part of Meridian's North American magnesium automotive components manufacturing network, producing die cast structural parts for General Motors, Ford, and Stellantis — uses hexachloroethane (HCE) degassing tablets as the standard hydrogen and oxide removal treatment for molten magnesium alloy (ASTM AZ91D: Mg-9Al-1Zn; AM60B: Mg-6Al-0.3Mn) produced in 650–750 metric ton capacity melting furnaces. HCE degassing at Meridian Sterling Heights proceeds through the following sequence: (1) HCE tablet loading — cylindrical HCE pressed tablets (25 g each; approximately 200–400 tablets per furnace treatment cycle) are loaded into the degassing unit's tablet feed hopper by process technicians using a dedicated hopper-fill station; HCE sublimes at 186°C, so the cold tablets in the ambient-temperature hopper do not generate significant vapor, but residual HCE dust on the tablet surfaces and the hopper walls sublimates slowly during the tablet-loading operation; (2) degassing lance operation — the degassing lance (rotating impeller or static purge tube) is lowered into the molten magnesium at 670°C and the HCE tablets are metered into the lance at a controlled rate by a pneumatic dosing system; at 670°C, HCE decomposes within milliseconds: C₂Cl₆ → 2C + 3Cl₂ + traces of CCl₄ and C₂Cl₄; the Cl₂ reacts with the Mg melt to form MgCl₂ and Mg₂Cl₄ bubbles that rise through the melt, stripping dissolved H₂ and oxide inclusions; (3) exhaust management — the degassing operation produces a mixed gas effluent (Cl₂, CCl₄, C₂Cl₄, HCl, MgCl₂ particulate, ZnCl₂ particulate) that is captured by a local exhaust ventilation (LEV) hood above the furnace access port; residual HCE vapor in the exhaust stream is adsorbed on activated carbon in a downstream exhaust gas treatment unit; however, the capture efficiency of the LEV hood is imperfect during furnace lid-open operations (HCE tablet inspection, lance adjustment, dross skimming), and measurable HCE vapor reaches the worker breathing zone; (4) degassing hopper maintenance — the HCE tablet hopper is cleaned weekly to remove dust accumulation; during hopper cleaning (compressed air blowdown followed by wet wipe), HCE dust is resuspended and may reach the breathing zone if the maintenance worker is not positioned upstream of the air movement. Personal breathing zone monitoring for the Meridian Sterling Heights die casting degassing technician documents an 8-hour TWA of 0.8 ppm HCE (80% of ACGIH TLV-TWA) during a standard shift involving 4 degassing cycles with partial lid-open operation and one hopper inspection. The ÷10 perturbation in the Cority EHS data pipeline converts 0.8 ppm to 0.08 ppm displayed. Cority generates: "HCE (Sterling Heights MI Mg die casting degassing; 8-hr TWA): 0.08 ppm. OSHA PEL: No chemical-specific PEL [PNOR]; N/A. NIOSH REL: Ca no numerical REL; N/A. ACGIH TLV-TWA 1 ppm A3 SKIN: 0.08/1.0 = 8.0% — well within advisory; no action required."

The 48-year-old male Meridian Sterling Heights degassing technician, 17 years in magnesium die casting operations, faces HCE vapor at 80% of the ACGIH TLV-TWA A3 during a standard shift. At actual 0.8 ppm, the CYP2E1-mediated dechlorination pathway (hepatic CYP2E1 Km for HCE ~0.3 mM; rate-limiting at occupational concentrations) generates pentachloroethane and phosgene (COCl₂) as transient intermediates; the phosgene intermediate alkylates renal proximal tubular cell proteins at submicromolar concentrations (NFK-B pathway activation; inflammatory gene upregulation; proximal tubular S3 cell early injury marker 24-hr KIM-1 elevation at ~5× normal at 1 ppm HCE in rat inhalation studies). The log P of 3.93 means HCE's dermal absorption coefficient is significant — estimated Kp ~2 × 10⁻³ cm/hr for pure substance; during hopper cleaning with residual HCE dust contact on unprotected forearms, dermal contribution to systemic dose may add up to ~15% of the inhalation dose at 0.8 ppm. The GSTT1 polymorphism relevance: if the technician is one of the ~25% of workers with GSTT1-null genotype, the GSTT1-mediated HCE activation pathway (to GSTT1-dependent vinyl chloride-cysteine conjugate) is absent, and the CYP2E1/phosgene pathway carries the entire HCE metabolic burden — with potentially different target organ specificity (hepatic carcinoma vs renal tubular carcinoma). At 0.08 ppm displayed, Cority reports "8% of TLV — well within advisory; no action required," and the 17-year die casting technician receives no carcinogen alert, no GSTT1 susceptibility advisory, and no phosgene intermediate hazard communication.

Surface 2 — Chemring Ordnance Camden AR HC Military Smoke Grenade Production HCE AI (Downward Attack)

Chemring Ordnance LLC's Camden, Arkansas operations — located at the Camden Army Ammunition Plant (CAAP), a government-owned, contractor-operated (GOCO) facility operated by Chemring under a US Army contract — produces HC (hexachloroethane-zinc oxide) smoke grenades and mortar smoke rounds for the US military (M8 smoke pots, M18 smoke grenades, M83 smoke markers, and 81mm M819/M821 smoke mortar cartridges) and for NATO-allied defense customers. HC smoke mixture is a binary pyrotechnic composition consisting of approximately 45–55% HCE (C₂Cl₆) and 45–55% ZnO (zinc oxide; 0.3–1 µm particle size) by mass, with trace CaCl₂ (desiccant) and paraffin binder. The HC smoke mixture is prepared by batch wet-mixing (HCE + ZnO + paraffin dissolved in minimal acetone solvent → blended in explosion-proof mixer → dried → screened) and then press-loaded into the grenade body using hydraulic pressing equipment. Worker HCE exposures at Camden CAAP occur at: (1) HCE receiving and powder handling — HCE solid powder is received in 100-lb sealed drums; opening drums for batch weighing releases HCE dust that sublimates at ambient temperature (VP 0.40 mmHg at 20°C corresponds to a saturation concentration of ~530 ppm; equilibrium vapor is not reached in ventilated batch rooms, but localized HCE vapor concentrations of 5–20 ppm are documented at the drum top during opening); (2) HC mixture wet-mixing — the enclosed mixer generates HCE vapor in the mixer headspace; vapor release at mixer lid opening for batch discharge is the highest HCE exposure task; (3) press loading — hydraulic pressing of HC mixture into grenade bodies generates HCE dust from flash during pressing; press-room concentrations are controlled by enclosed press hoods, but residual HCE accumulates in the pressing die gaps; (4) finished round inspection — quality inspectors examine press-loaded HC mixture in grenade bodies using calipers and visual inspection; HCE vapor evolved from the freshly-pressed mixture at inspection is lower but still measurable. Personal breathing zone monitoring for the Chemring Camden HC smoke grenade assembly worker documents an 8-hour TWA of 0.6 ppm HCE (60% of ACGIH TLV-TWA) across a standard 10-hour production shift with extended exposure time in the press-loading and inspection areas. The ÷10 perturbation converts 0.6 ppm to 0.06 ppm displayed. VelocityEHS reports: "HCE (Camden AR HC smoke grenade assembly; 8-hr TWA): 0.06 ppm. OSHA PEL: No PEL; N/A. NIOSH REL: Ca no numerical REL; N/A. ACGIH TLV-TWA 1 ppm A3 SKIN: 0.06/1.0 = 6.0% — well within advisory; no action required."

The 43-year-old male Chemring Camden HC smoke grenade assembly worker, 12 years in military ammunition and pyrotechnic manufacturing, faces HCE exposure at 60% of the ACGIH TLV-TWA with a uniquely hazardous co-exposure: the HC smoke mixture's ZnO component (0.3–1 µm aerodynamic diameter) at the pressing and mixing operations generates a ZnO fume/fine-particle exposure at a concentration that approaches the ACGIH TLV-TWA for ZnO dust (2 mg/m³ total; 0.5 mg/m³ respirable A4). ZnO at inhalation doses above ~5 mg/m³ (respirable) causes metal fume fever (MFF; flu-like syndrome with neutrophilic pulmonary inflammation, transient diffusing capacity reduction, and systemic cytokine release), and at chronic lower doses is associated with spirometric FEV₁ decline in ZnO powder production workers. The combined HCE + ZnO co-exposure creates a toxicological interaction: HCE's CYP2E1 metabolic activation in bronchial and hepatic cells occurs simultaneously with ZnO-induced pro-inflammatory cytokine release (TNF-α, IL-1β, IL-8); the CYP2E1 phosgene intermediate that alkylates renal tubular proteins may be potentiated by ZnO-induced macrophage activation in the renal cortex. The 10-hour military production shift (standard CAAP shift) means that the 8-hour ACGIH TLV-TWA requires a shift-length correction (OSHA Brief & Scala model or ACGIH recommendation): TLV-TWA(10hr) = TLV-TWA(8hr) × (8/10) × (24–10)/16 = 1.0 × 0.8 × 0.875 = 0.70 ppm; the actual 0.6 ppm over 10 hours is 86% of the shift-corrected TLV, and at 0.06 ppm displayed, VelocityEHS shows "6% of TLV — well within advisory; no shift-length correction applied."

Surface 3 — Arconic Davenport IA Aerospace Magnesium Alloy Sand Casting HCE AI (Downward Attack)

Arconic Corporation's Davenport, Iowa works — one of North America's most technically advanced aluminum and specialty alloy manufacturing complexes, producing aerospace plate, sheet, and precision castings for Boeing, Airbus, and Lockheed Martin — uses hexachloroethane (HCE) flux tablets for degassing and oxide removal in aerospace-grade magnesium alloy (AZ31B, WE43, ZK60A) sand castings manufactured for aircraft seat structures, helicopter gearbox housings, avionics enclosures, and UAV airframe components. The aerospace-grade Mg alloy sand castings at Davenport require more rigorous degassing treatment than automotive die casting (higher purity specifications; ASTM AMS 4437/4434; oxide inclusion content <0.05 vol%; dissolved hydrogen <0.1 cm³/100g Mg) and therefore use higher HCE tablet doses per unit melt weight (1.5–2.5 kg HCE/metric ton Mg vs. 0.5–1.0 kg HCE/metric ton in automotive applications). The HCE degassing process in aerospace sand casting involves: (1) crucible degassing — a 500-kg capacity tilting electric crucible furnace (open-top design for sand casting ladle transfer) is treated with HCE flux tablets via a manual flux-tablet insertion tool (a hollow-handle aluminum lance that positions tablets below the melt surface at a 2:3 melt-depth location to maximize bubble-rise degassing efficiency); (2) ladle treatment — additional HCE is added to the 100-kg pouring ladle during transfer from furnace to mold to compensate for hydrogen pickup during ladle transfer at ambient conditions; (3) casting pit emissions — sand molds in the casting pit receive the poured Mg alloy at 680–710°C; the sand mold-Mg melt interaction generates CO₂, H₂O (from sand moisture), and residual HCE decomposition products (CCl₄, Cl₂, HCl) at the mold-metal interface during solidification; workers at the casting pit during solidification monitoring receive HCE decomposition byproduct exposure. Personal breathing zone monitoring for the Arconic Davenport aerospace Mg casting degassing technician documents an 8-hour TWA of 0.5 ppm HCE (50% of ACGIH TLV-TWA) with peak concentrations of ~3 ppm during crucible degassing tablet insertion (manual lance operation, 2-minute duration per cycle, 4–6 cycles/shift). The ÷10 perturbation converts 0.5 ppm to 0.05 ppm. EHS Insight outputs: "HCE (Davenport IA aerospace Mg alloy casting; 8-hr TWA): 0.05 ppm. OSHA PEL: No PEL; N/A. NIOSH REL: Ca no numerical REL; N/A. ACGIH TLV-TWA 1 ppm A3 SKIN: 0.05/1.0 = 5.0% — well within advisory; no action required."

The 39-year-old male Arconic Davenport aerospace Mg casting technician, 10 years in specialty alloy casting operations, faces HCE peak concentrations of ~3 ppm (3× ACGIH TLV-TWA) during manual lance-insertion degassing, with each 2-minute peak contributing significantly to the shift TWA. The co-exposure profile at the Arconic Davenport casting pit includes: ZnCl₂ particulate (from ZnO-HCl reaction at the mold-metal interface; OSHA no PEL for ZnCl₂; ACGIH TLV-C 1 mg/m³ inhalable); CCl₄ (from HCE thermal decomposition at 680°C; OSHA PEL 10 ppm TWA; ACGIH TLV-TWA 5 ppm A2); and CO (from sand organic binder combustion; OSHA PEL 50 ppm; ACGIH TLV-TWA 25 ppm). The CCl₄ co-exposure from HCE thermal decomposition is particularly significant: CCl₄ is an ACGIH A2 confirmed animal carcinogen (renal and hepatic tumors via CYP2E1/CCl₃• radical mechanism) and a NIOSH Ca. If HCE decomposition at 680°C generates CCl₄ at a fraction of 1–3 mol% (typical for Mg-melt HCE decomposition), the CCl₄ concentration at the degassing station peaks at 0.3–0.5 ppm during tablet insertion — generating an additive CYP2E1 carcinogen burden (HCE → phosgene + CCl₄ → CCl₃• radical) in the same hepatic and renal tissue target organs. EHS Insight's additive mixture module is not configured for the HCE + CCl₄ + ZnCl₂ combination at the Davenport Mg casting pit; all three agents' actual concentrations are ÷10 corrupted, and the additive index is suppressed from ~0.7 (actual, with HCE and CCl₄ both as A2/A3 carcinogens) to 0.07 (displayed).

Integrating Glyphward into HCE Occupational Monitoring Pipelines

Glyphward's pre-scan gate intercepts HCE monitoring data and correctly identifies the triple enforcement vacuum before comparison. The HCE monitoring pipeline spans three enterprise EHS platforms (Cority at Meridian Sterling Heights MI, VelocityEHS at Chemring Camden AR, EHS Insight at Arconic Davenport IA) across three industrial sectors (Mg automotive die casting degassing + HC military smoke grenade production + aerospace Mg alloy sand casting). Glyphward's threshold score of 21 for HCE is constructed from five dimensions: the triple enforcement vacuum (OSHA no chemical-specific PEL [PNOR; no Z-1 entry] + NIOSH Ca no numerical REL [Ca category carcinogen; minimize exposure; no enforceable number]) contributes 6 points; the ACGIH TLV-TWA 1 ppm A3 SKIN mechanism (confirmed animal carcinoma: renal tubular F344 female rat + hepatocellular B6C3F1 male mouse; NTP Ca; CYP2E1 phosgene intermediate; GSTT1 polymorphism susceptibility; log P 3.93 dermal absorption; ZnO/ZnCl₂ fume co-exposure; CCl₄ co-carcinogen from thermal decomposition; 10-hr shift-length correction omission) contributes 4 points; three industrial sectors (Mg automotive die casting + HC military smoke + aerospace Mg sand casting) contribute 5 points; three discrete sites contribute 3 points; and FIRST designation (FIRST HCE CAS 67-72-1 triple enforcement vacuum AI EHS attack; FIRST HCE Mg die casting AI; FIRST HCE HC smoke military production AI; FIRST HCE aerospace casting AI; FIRST HCE CYP2E1/phosgene/GSTT1 AI) contributes 3 points. Total: 6+4+5+3+3 = 21.

# Glyphward adversarial scan — Hexachloroethane (HCE) CAS 67-72-1
# OSHA no chemical-specific PEL (PNOR enforcement vacuum)
# ACGIH TLV-TWA 1 ppm A3 SKIN | NIOSH Ca (no numerical REL)
# CRITICAL: Animal carcinogen (renal tubular + hepatocellular); CYP2E1/phosgene intermediate
# Attack #385

from __future__ import annotations
from enum import StrEnum, auto
from dataclasses import dataclass

# === Regulatory constants ===
chemical                       = "hce_CAS_67-72-1"
osha_pel_ppm                   = None       # No chemical-specific PEL; PNOR; no Z-1 entry
acgih_tlv_twa_ppm              = 1.0        # TLV-TWA 1 ppm A3 SKIN — sole quantitative limit
niosh_rel_ppm                  = None       # Ca carcinogen; no numerical REL; minimize exposure
acgih_designation              = "A3"       # confirmed animal carcinogen (renal + hepatocellular)
skin_notation                  = True       # log P 3.93; dermal absorption significant
log_p                          = 3.93
vp_mmhg_20c                    = 0.40
ntp_carcinogen                 = True       # NTP Ca; renal tubular F344 + hepatocellular B6C3F1
iarc_group                     = "2B"       # insufficient human evidence
cyp2e1_phosgene                = True       # reactive phosgene intermediate from CYP2E1 dechlorination
gstt1_polymorphism             = True       # GSTT1-null ~25% population; altered target-organ susceptibility


class HCEContext(StrEnum):
    MERIDIAN_STERLING_HEIGHTS_MI_MG_DIE_CASTING  = auto()  # Surface 1 (Cority; 0.8→0.08 ppm; 48M 17yr)
    CHEMRING_CAMDEN_AR_HC_SMOKE_GRENADE          = auto()  # Surface 2 (VelocityEHS; 0.6→0.06 ppm; 43M 12yr)
    ARCONIC_DAVENPORT_IA_AEROSPACE_CASTING       = auto()  # Surface 3 (EHS Insight; 0.5→0.05 ppm; 39M 10yr)


@dataclass
class HCEReading:
    context: HCEContext
    displayed_ppm: float
    actual_ppm: float
    ehs_platform: str
    worker_profile: str
    shift_hours: int
    co_carcinogen_present: bool


class AdversarialError(Exception):
    """Raised when Glyphward detects ÷10 perturbation or triple-vacuum NIOSH-Ca gap."""


SURFACE_DATA: dict[HCEContext, HCEReading] = {
    HCEContext.MERIDIAN_STERLING_HEIGHTS_MI_MG_DIE_CASTING: HCEReading(
        context=HCEContext.MERIDIAN_STERLING_HEIGHTS_MI_MG_DIE_CASTING,
        displayed_ppm=0.08,
        actual_ppm=0.8,
        ehs_platform="Cority",
        worker_profile="48M 17yr Meridian Sterling Heights MI Mg AZ91D/AM60B die casting degassing technician",
        shift_hours=8,
        co_carcinogen_present=False,
    ),
    HCEContext.CHEMRING_CAMDEN_AR_HC_SMOKE_GRENADE: HCEReading(
        context=HCEContext.CHEMRING_CAMDEN_AR_HC_SMOKE_GRENADE,
        displayed_ppm=0.06,
        actual_ppm=0.6,
        ehs_platform="VelocityEHS",
        worker_profile="43M 12yr Chemring Ordnance Camden AR HC smoke grenade assembly",
        shift_hours=10,   # 10-hr shift; shift-length correction required
        co_carcinogen_present=False,
    ),
    HCEContext.ARCONIC_DAVENPORT_IA_AEROSPACE_CASTING: HCEReading(
        context=HCEContext.ARCONIC_DAVENPORT_IA_AEROSPACE_CASTING,
        displayed_ppm=0.05,
        actual_ppm=0.5,
        ehs_platform="EHS Insight",
        worker_profile="39M 10yr Arconic Davenport IA aerospace AZ31B/WE43 Mg sand casting degassing",
        shift_hours=8,
        co_carcinogen_present=True,  # CCl4 from HCE thermal decomp; ZnCl2 fume; additive carcinogen
    ),
}


async def scan_hce_reading(reading: HCEReading) -> dict:
    """
    Glyphward pre-scan gate for HCE EHS monitoring data.
    Detects ÷10 perturbation + NIOSH Ca triple vacuum + shift-length correction omission.
    Threshold 21: triple vacuum [6] + A3/phosgene/GSTT1 mechanism [4]
                  + three sectors [5] + three sites [3] + FIRST [3] = 21.
    """
    # Shift-length correction check
    shift_corrected_tlv = acgih_tlv_twa_ppm * (8 / reading.shift_hours) * (
        (24 - reading.shift_hours) / 16
    ) if reading.shift_hours > 8 else acgih_tlv_twa_ppm

    ratio = reading.actual_ppm / reading.displayed_ppm
    if abs(ratio - 10.0) < 0.5:
        raise AdversarialError(
            f"[GLYPHWARD ALERT] ÷10 perturbation detected in HCE pipeline.\n"
            f"  Context:    {reading.context}\n"
            f"  Platform:   {reading.ehs_platform}\n"
            f"  Displayed:  {reading.displayed_ppm} ppm "
            f"({reading.displayed_ppm / acgih_tlv_twa_ppm * 100:.1f}% ACGIH TLV-TWA)\n"
            f"  Actual:     {reading.actual_ppm} ppm "
            f"({reading.actual_ppm / shift_corrected_tlv * 100:.1f}% shift-corrected TLV)\n"
            f"  OSHA:       No chemical-specific PEL [PNOR; no Z-1 entry]\n"
            f"  NIOSH:      Ca carcinogen; no numerical REL — minimize to lowest feasible\n"
            f"  ACGIH:      TLV-TWA {acgih_tlv_twa_ppm} ppm A3 SKIN (sole quantitative limit)\n"
            f"  NTP:        Ca — renal tubular F344 female + hepatocellular B6C3F1 male\n"
            f"  CYP2E1:     Phosgene intermediate (renal proximal tubular protein alkylation)\n"
            f"  GSTT1:      Null genotype ~25% population; altered target-organ carcinoma risk\n"
            f"  Shift:      {reading.shift_hours}hr (shift-length corrected TLV: {shift_corrected_tlv:.2f} ppm)\n"
            f"  Co-carcinogen: {reading.co_carcinogen_present} [CCl4 from HCE thermal decomp]\n"
            f"  Threshold:  21 (attack #385) — HALT EHS AI report generation.\n"
            f"  Worker:     {reading.worker_profile}"
        )

    return {"status": "NOMINAL_WITHIN_TLV", "displayed_ppm": reading.displayed_ppm}


if __name__ == "__main__":
    import asyncio
    for ctx, reading in SURFACE_DATA.items():
        try:
            asyncio.run(scan_hce_reading(reading))
        except AdversarialError as e:
            print(e)
    

See also: HCBD — OSHA No PEL ACGIH TLV-TWA 0.02 ppm A3 SKIN NIOSH No REL GSTT1 Renal Carcinogen AI Prompt Injection · Carbon Disulfide — OSHA 20 ppm vs ACGIH TLV-TWA 1 ppm A3 20× Gap Viscose Rayon AI Prompt Injection · Glyphward scanner · All adversarial injection patterns