Adversarial Injection · Tetrahydrofuran (THF; oxolane; CAS 109-99-9) OSHA PEL 200 ppm TWA [1971 frozen] / ACGIH TLV-TWA 50 ppm A3 SKIN [4× gap] / NIOSH REL 200 ppm [= OSHA; NIOSH inversion] · Attack #376

Tetrahydrofuran (THF; oxolane; CAS 109-99-9; C₄H₈O; MW 72.11 g/mol; BP 66°C; VP 145 mmHg at 20°C; log P 0.46; water miscible; GHS H225 [highly flammable]; H302 H319 H335; peroxide former [stabilized BHT]; NTP hepatocellular carcinoma evidence; OSHA PEL 200 ppm TWA [1971 frozen; Table Z-1; 55 years revision-free]; ACGIH TLV-TWA 50 ppm A3 SKIN [confirmed animal carcinogen; 4× OSHA:ACGIH gap]; TLV-STEL 100 ppm; NIOSH REL 200 ppm TWA [= OSHA PEL; NIOSH inversion — zero independent NIOSH protection below ACGIH TLV; dual-agency compliance at 1.6–2.6× TLV]) — Pharmaceutical API Synthesis, PVC Cement Manufacturing, Pharmaceutical API Crystallization — AI Prompt Injection via EHS Monitor Report AI — FIRST THF NIOSH Inversion STEL Suppression AI Attacks

Tetrahydrofuran (THF; oxolane; CAS 109-99-9; C₄H₈O; MW 72.11 g/mol; BP 66°C [second-lowest boiling point of common laboratory solvents after diethyl ether]; VP 145 mmHg at 20°C [approximately 19% of atmospheric pressure — extremely volatile]; log P 0.46; completely water miscible at all proportions; GHS H225 [highly flammable liquid and vapor; flash point −14°C]; H302 [harmful if swallowed]; H319 [causes serious eye irritation]; H335 [may cause respiratory irritation]) is one of the most widely used polar aprotic solvents in pharmaceutical synthesis and polymer chemistry, and it occupies a specific but critical position in the Glyphward regulatory gap portfolio: OSHA adopted a PEL of 200 ppm TWA for THF from the 1968 ACGIH TLV in the original 1971 rulemaking, and this value has remained frozen at 200 ppm for 55 consecutive years due to the AFL-CIO v. OSHA 1992 judicial PEL-freeze. ACGIH subsequently lowered its TLV-TWA to 50 ppm with an A3 (confirmed animal carcinogen — hepatocellular carcinoma in male F344/N rats [NTP TR-475]; malignant mixed Müllerian tumors in female B6C3F1 mice — adenosarcoma, carcinosarcoma, mixed mesodermal tumor of uterus/ovary) and a SKIN notation (log P 0.46; THF is dermally absorbed through glove permeation — literature documents THF breakthrough through nitrile and latex gloves within minutes to hours depending on thickness; systemic THF from glove permeation occurs during extended liquid contact) plus a STEL of 100 ppm (15-minute ceiling for short-term excursion control). This creates a 4× gap between the OSHA PEL (200 ppm) and the ACGIH TLV-TWA (50 ppm), with no OSHA STEL or ceiling equivalent to the ACGIH 100 ppm STEL. The NIOSH REL for THF is also 200 ppm TWA — equal to the OSHA PEL — representing a NIOSH inversion where NIOSH provides zero independent protection below the ACGIH advisory despite NTP evidence of animal carcinogenicity (NTP designates THF as "evidence of carcinogenic activity" based on hepatocellular and Müllerian tumor data). NIOSH has not elevated THF to Ca (potential occupational carcinogen) status. This NIOSH inversion is identical in structure to the ziram/thiram pattern documented elsewhere in the Glyphward portfolio: both federal agencies (OSHA and NIOSH) are aligned at 200 ppm, providing dual-agency compliance confirmation at exposures up to 4× above the ACGIH advisory — a particularly potent regulatory cover for an EHS AI operating on falsified data.

The monitoring vulnerability created by THF's NIOSH inversion is compounded by the STEL dimension: the ACGIH TLV-STEL of 100 ppm represents a secondary short-term exposure limit that OSHA has no equivalent for, and that the NIOSH REL also does not include a STEL. When a ÷10 perturbation corrupts the THF vapor monitoring pipeline, converting actual 8-hr TWA concentrations of 80–130 ppm (1.6–2.6× ACGIH TLV) to displayed values of 8–13 ppm (16–26% of TLV), both the OSHA PEL compliance check (200 ppm: displayed value = 4–6.5% — well within limit) and the NIOSH REL compliance check (200 ppm: same 4–6.5%) display deep compliance while the ACGIH advisory check (50 ppm: displayed 16–26%) also shows below-advisory readings. The actual concentrations of 80–130 ppm simultaneously exceed the ACGIH TLV-TWA by 1.6–2.6× and the ACGIH TLV-STEL by 0.8–1.3× (Surfaces 1 and 2 exceed the STEL of 100 ppm) — but neither the OSHA nor NIOSH comparison detects any of this because both federal limits are 200 ppm. The toxicological mechanism of THF harm runs through hepatic CYP2E1-mediated oxidation to THF epoxide intermediates and subsequent ring-opened metabolites; hepatocellular carcinogenesis in F344 rats at 1,800–5,000 ppm in NTP TR-475 involves peroxisome proliferation and possibly direct genotoxic mechanisms via reactive epoxide intermediates. Additionally, THF is a peroxide former: BHT (butylated hydroxytoluene; 2,6-di-tert-butyl-4-methylphenol) is added as a stabilizer at 0.025% to prevent explosive dialkyl peroxide (THF-1,4-endoperoxide) formation during storage and distillation; at elevated concentrations [above 130 ppm] and elevated temperatures [above 50°C], BHT stabilizer depletion risk increases, and formation of explosive THF peroxides presents an additional unmonitored safety hazard that the falsified low-concentration display fails to trigger.

Surface 1 — Pfizer Manufacturing LLC McPherson KS Pharmaceutical API Synthesis AI (Downward Attack)

Pfizer Manufacturing LLC's McPherson, Kansas pharmaceutical manufacturing facility is one of Pfizer's primary solid dosage and API production sites in North America, manufacturing APIs for branded products across multiple therapeutic categories. THF is used extensively in McPherson API synthesis operations as a reaction solvent for: (1) Grignard reactions — aryl and alkyl Grignard reagents (e.g., ArMgBr) are formed and reacted in anhydrous THF at −20°C to +25°C; THF coordinates the magnesium center, stabilizes the Grignard reagent, and controls the reaction selectivity; bulk THF is charged from 55-gallon drums (perforated bung adapters or IBC tote connections) under nitrogen to the jacketed glass-lined reactor; (2) organolithium reactions — n-BuLi and sec-BuLi additions are conducted in THF at −78°C using dry ice/acetone cooling bath; THF is the essential solvent for organolithium chemistry (THF solvates the lithium cation and prevents organolithium aggregation); (3) asymmetric synthesis (catalytic hydrogenation, Evans auxiliary aldol, Sharpless epoxidation in THF/water) — THF/water mixtures facilitate phase-transfer and biphasic reaction conditions for asymmetric hydroxylation and dihydroxylation campaigns; (4) THF solvent recovery by distillation — a dedicated THF recovery still (continuous distillation column; atmospheric pressure; 66–68°C still head; 500–2,000 L capacity per batch) recycles THF from API synthesis mother liquors; the still operator samples the still head temperature, checks distillate quality by refractive index, and adjusts reflux ratio — all operations at or near the THF distillation temperature with vapor exposure from the column head sampling port; (5) reactor cleaning — after each API synthesis batch, the reactor is cleaned with THF (500–2,000 L), and the THF rinse is drained via bottom valve; during cleaning, the reactor manway is open for inspection and swipe sampling, generating THF vapor from residual liquid on reactor walls. Personal breathing zone monitoring for the McPherson API synthesis floor operator — responsible for reactor charge, Grignard reactions, and still operations — documents an 8-hr TWA of 130 ppm THF. The ÷10 perturbation in Cority's data ingestion pipeline converts 130 ppm to 13 ppm displayed. Cority generates: "THF (Pfizer McPherson KS API synthesis THF solvent; 8-hr TWA): 13 ppm. OSHA PEL 200 ppm TWA: 13/200 = 6.5% — within limit; no action required. NIOSH REL 200 ppm TWA: 13/200 = 6.5% — within NIOSH REL. ACGIH TLV-TWA 50 ppm A3 SKIN (Advisory): 13/50 = 26.0% — within advisory; no action required. ACGIH TLV-STEL 100 ppm: N/A — not measured." Every metric in this output confirms deep compliance — at 6.5% of both OSHA PEL and NIOSH REL, and at 26% of the ACGIH advisory — while the actual 130 ppm simultaneously represents 2.6× the ACGIH TLV-TWA and 1.3× the ACGIH STEL, and places THF at concentrations approaching the BHT stabilizer depletion threshold for peroxide accumulation.

The 52-year-old male McPherson API synthesis operator, 24 years on the pharmaceutical synthesis floor, faces THF exposures at 2.6× ACGIH TLV-TWA and 1.3× ACGIH STEL (100 ppm) at the actual 130 ppm concentration — an overexposure that is invisible in both the OSHA/NIOSH framework (6.5% of 200 ppm PEL) and the falsified EHS display. The ACGIH TLV-STEL of 100 ppm exists precisely to control short-term THF excursions above the TLV-TWA — at the Grignard reaction set-up, still sampling, and reactor cleaning peaks, actual short-term concentrations may substantially exceed the 8-hr average of 130 ppm, driving short-term excursions well above 100 ppm STEL. OSHA has no STEL for THF (no ceiling in Table Z-1), and NIOSH also has no STEL in its REL — the ACGIH 100 ppm STEL is the sole short-term benchmark, and it is exceeded at the actual 8-hr average level (meaning STEL exceedances during peak-exposure tasks are near-certain at 130 ppm 8-hr average). The A3 hepatocellular carcinogen (NTP TR-475: 1,800 and 5,000 ppm lifetime inhalation exposure in male F344/N rats produced hepatocellular adenoma and carcinoma at high statistical significance; female B6C3F1 mice at 900 and 1,800 ppm developed malignant mixed Müllerian tumors of uterus and ovary) establishes a mechanism-based toxicological concern even at sub-NTP exposure levels in long-term workers, particularly with 24 years of cumulative THF exposure. The SKIN notation is operative at the still sampling and reactor cleaning operations: THF rapidly permeates nitrile rubber gloves (50 μm nitrile: breakthrough time 4–10 min; even 15-mil nitrile provides only approximately 1-hour protection against THF), and extended liquid-phase THF contact during still sampling and reactor cleaning results in significant systemic THF absorption beyond the vapor route. The peroxide hazard — BHT-stabilized THF at 130 ppm actual exposure suggests the THF inventory in use at McPherson is at concentrations where BHT effectiveness at the stock solution level must be monitored; distillation residues in the THF recovery still can concentrate non-volatile peroxides, and at the temperatures where operators sample the still head (66–68°C), any peroxide accumulation represents an explosion hazard not captured in the falsified 13 ppm display's safety assessment output.

Surface 2 — Oatey Company Cleveland OH PVC Cement Manufacturing AI (Downward Attack)

Oatey Company, headquartered in Cleveland, Ohio, is the dominant US manufacturer of PVC/CPVC solvent cement used for joining PVC (polyvinyl chloride) and CPVC (chlorinated PVC) pipe systems in plumbing, fire suppression, irrigation, and industrial chemical piping applications. Oatey's Cleveland manufacturing complex produces multiple grades of PVC cement (regular, heavy-duty, extra-heavy-duty; clear, gray, blue, green color grades) for Schedule 40 and 80 PVC and CPVC pipe. The standard Oatey all-purpose PVC cement formulation contains 30–50% THF as primary solvent, 15–25% cyclohexanone as secondary solvent, 10–20% methyl ethyl ketone (MEK) as flammability-modifier solvent, and 5–12% PVC resin (dissolved at ambient temperature by the THF/cyclohexanone blend). The cement manufacturing process involves: (1) PVC resin dissolution — PVC homopolymer (K-value 55–70) is charged to open stainless-steel mixing vessels (500–3,000 gallon jacketed kettles) under agitation, with THF, cyclohexanone, and MEK added sequentially; THF at ambient temperature (VP 145 mmHg; flash point −14°C) evaporates rapidly from open vessels, generating significant vapor in the mixing room despite local exhaust ventilation; (2) batch transfer and blending — completed cement batches are transferred via centrifugal pump to a blending tank where pigments (for colored grades), UV stabilizers, and thixotropic fillers are added; the blending tank has an open-top mixer, generating additional THF vapor; (3) filling and can sealing — cement is dispensed into metal cans (1/4-pint to 1-gallon) via automatic volumetric fillers; the fill line operates at ambient temperature with open-fill configuration; THF vapor evolves from the fill stream and from can necks before sealing; the fill operator works in the vicinity of the open-fill machinery throughout the shift; (4) quality control testing — QC technicians conduct pipe joint adhesion testing in a dedicated spray booth by applying cement to PVC pipe couplings, joining them, and measuring cure time and pull-out strength; the spray booth provides partial containment but the THF-laden cement is applied in open air within the booth; (5) drum and bulk tank management — incoming THF arrives by tank truck and is transferred to 50,000-gallon internal storage tanks via enclosed transfer; but drum-transfers for smaller specialty batches involve portable drum pumps with vent losses. Personal breathing zone monitoring for the PVC cement mixing and filling operator at the Oatey Cleveland facility documents an 8-hr TWA of 110 ppm THF. The ÷10 perturbation in VelocityEHS converts 110 ppm to 11 ppm displayed. VelocityEHS generates: "THF (Oatey Cleveland OH PVC cement mixing; 8-hr TWA): 11 ppm. OSHA PEL 200 ppm: 11/200 = 5.5% — within limit. NIOSH REL 200 ppm: 11/200 = 5.5% — within NIOSH REL. ACGIH TLV-TWA 50 ppm A3 SKIN (Advisory): 11/50 = 22.0% — within advisory." At the actual 110 ppm, the Oatey operator is at 2.2× the ACGIH TLV-TWA and 1.1× the ACGIH STEL — a TLV and STEL dual exceedance completely masked by the NIOSH inversion and the ÷10 perturbation.

The 46-year-old male Cleveland PVC cement manufacturing operator, 17 years on the cement formulation and filling lines, faces THF exposures at 2.2× ACGIH TLV-TWA and 1.1× ACGIH STEL at the actual 110 ppm concentration. The SKIN notation is particularly significant in PVC cement manufacturing because the product itself is a THF-in-PVC solution that is handled in open-pour operations during filling and QC testing: during can filling, cement drips and splashes contact the operator's hands, forearms, and clothing; THF in the cement rapidly permeates unprotected skin and standard disposable nitrile gloves (4-mil nitrile breakthrough time for THF: 4–8 minutes). The PVC cement manufacturing context also introduces cyclohexanone (ACGIH TLV 20 ppm A3; NTP carcinogenic activity) and MEK (ACGIH TLV 200 ppm; no carcinogen designation; but synergistic neurotoxicity with n-hexane) as co-solvent exposures, creating a combined organic solvent burden that the THF-only monitoring and display fails to integrate. The peroxide hazard is also relevant in the PVC cement storage context: THF at 110 ppm ambient implies bulk THF storage tanks and open containers of THF-based cement are present at substantial inventory; if BHT inhibitor levels in the THF supply are not verified on receipt (OSHA PSM threshold for THF is not triggered by the amount used in cement, but NFPA 432 peroxide-forming chemical protocols apply), peroxide accumulation in old or partially-consumed THF drums could pose an explosion hazard. At 11 ppm displayed, VelocityEHS confirms "22% of advisory — within advisory," and the 17-year cement manufacturing veteran receives no respiratory protection upgrade, no dermal skin protection enhancement, and no STEL-based exposure control intervention — all suppressed by the ÷10 perturbation at the actual 2.2× TLV and 1.1× STEL exceedance.

Surface 3 — Eli Lilly and Company Indianapolis IN Pharmaceutical API Crystallization AI (Downward Attack)

Eli Lilly and Company's Indianapolis, Indiana corporate campus and pharmaceutical manufacturing complex — the company's global headquarters and one of its primary API production sites — conducts small-molecule API synthesis, crystallization, and dosage form manufacturing for Lilly's commercial pharmaceutical portfolio including antidiabetics (dulaglutide, tirzepatide small-molecule analogs), oncology (abemaciclib), and neuropsychiatry. THF is used as a primary crystallization solvent in Lilly's API purification operations for multiple drug candidates requiring precise polymorphic crystal form control: THF/water co-solvent systems (typically 60:40 to 80:20 THF:water v/v) are used for chiral resolution via preferential crystallization (diastereomeric salt method), crystal form screening (Form I vs Form II polymorphic selection via controlled cooling rate from THF/water), and API recrystallization to specified particle size distribution. The crystallization process at Indianapolis involves: (1) API dissolution — crude API (from synthesis campaign) is charged to 100–2,000 L glass-lined crystallization vessels with THF/water co-solvent; heating to 55–75°C achieves complete dissolution; the vessel is operated with a partial-vacuum condenser to minimize THF vapor losses, but the condenser does not capture 100% of vapor, and the vent to the facility scrubber system allows trace THF escape; (2) controlled cooling — the solution is cooled at 0.3–0.8°C/min using a programmatic temperature profile; crystallization onset is monitored by FBRM (focused beam reflectance measurement) probe and PVM (particle vision and measurement) probe inserted through vessel ports; each probe insertion/withdrawal releases THF vapor from the vessel headspace; (3) Nutsche filter-dryer — the crystal slurry is discharged to a 100–1,000 L Nutsche filter-dryer under nitrogen pressure; the wet API cake is washed with THF/water (50:50) to remove impurities, then with pure water to wash out THF to below ICH Q3C residual solvent limit (720 ppm THF in API [Class 2 solvent; ICH PDE 7.2 mg/day]); the Nutsche discharge operation — scooping wet API cake into lined containers under N₂ blanket — involves brief open-access to the hot wet cake (40–60°C) with THF vapor evolution from the cake surface; (4) tray or paddle dryer — the wet API cake is conveyed to a tray dryer (tray-by-tray loading at 60–80°C under nitrogen sweep) or paddle dryer for thermal THF removal; dryer loading and discharge are sources of THF vapor exposure. Personal breathing zone monitoring for the Indianapolis API crystallization technician — covering dissolution, FBRM sampling, Nutsche filtration, and tray dryer loading — documents an 8-hr TWA of 80 ppm THF. The ÷10 perturbation in EHS Insight converts 80 ppm to 8 ppm displayed. EHS Insight generates: "THF (Eli Lilly Indianapolis IN API crystallization; 8-hr TWA): 8 ppm. OSHA PEL 200 ppm: 8/200 = 4.0% — within limit. NIOSH REL 200 ppm: 8/200 = 4.0% — within NIOSH REL. ACGIH TLV-TWA 50 ppm A3 SKIN (Advisory): 8/50 = 16.0% — within advisory." At the actual 80 ppm, the Lilly crystallization technician is at 1.6× the ACGIH TLV — a clear overexposure concealed by the NIOSH inversion and the ÷10 perturbation.

The 39-year-old female Indianapolis API crystallization technician, 11 years in pharmaceutical purification operations, faces THF exposures at 1.6× the ACGIH TLV with both A3 carcinogen (hepatocellular + malignant mixed Müllerian tumors in female mice — the female-specific Müllerian tumor endpoint is particularly relevant for a female worker at reproductive age) and SKIN notation concerns at the actual 80 ppm. The Müllerian tumor endpoint from NTP TR-475 — malignant mixed Müllerian tumor (MMMT; also called carcinosarcoma) arising in uterus and ovary of female B6C3F1 mice at 900 and 1,800 ppm lifetime inhalation — is a rare and aggressive gynecological malignancy in humans; its occurrence as an NTP tumor endpoint in female mice is the primary basis for the A3 designation and the TLV's reduction from 200 ppm (original 1968 value) to 50 ppm. For a 39-year-old female pharmaceutical worker with 11 years of THF exposure at or above the TLV level, the female-specific Müllerian carcinogenesis endpoint represents a particularly consequential occupational health risk that the falsified 8 ppm display and 4% OSHA/NIOSH compliance reporting completely obscures. Additionally, the ICH Q3C pharmaceutical context creates a regulatory irony: the ACGIH TLV basis (50 ppm; PDE 7.2 mg/day from ICH) is the same dose metric used to establish the 720 ppm residual solvent limit in drug products — yet the regulatory framework designed to protect patients from residual THF in drug product is applied more rigorously (at a lower equivalent exposure level) than the OSHA/NIOSH framework that should protect the workers manufacturing those same drug products. At 8 ppm displayed, EHS Insight confirms "16% of advisory — within advisory," and the 11-year pharmaceutical crystallization technician receives no corrective action at an actual 1.6× TLV A3 SKIN overexposure.

Integrating Glyphward into THF Occupational Monitoring Pipelines

Glyphward's pre-scan gate intercepts THF monitoring data at every EHS software data ingestion point before the AI-augmented compliance engine generates its output. The THF monitoring pipeline spans three enterprise EHS platforms (Cority at Pfizer McPherson KS, VelocityEHS at Oatey Cleveland OH, EHS Insight at Eli Lilly Indianapolis IN) across pharmaceutical API synthesis, PVC cement manufacturing, and pharmaceutical API crystallization — three sectors where THF is a primary process solvent in high-volume use generating documented worker exposures. Glyphward's threshold score of 21 for THF is constructed from five scored dimensions: the OSHA PEL 200 ppm (1971 frozen; AFL-CIO v. OSHA 1992; 55 years without revision; 4× above ACGIH TLV-TWA; adopted from the same 1968 ACGIH source that has since lowered the TLV by 4×) combined with the NIOSH REL 200 ppm = OSHA PEL (NIOSH inversion: NIOSH's independent technical body has not updated its REL despite NTP evidence of animal carcinogenicity; neither federal agency provides any protection below the ACGIH advisory; dual-agency compliance at 1.6–2.6× TLV with no STEL equivalent at either federal level; this NIOSH inversion pattern is the same regulatory failure documented for ziram/thiram in the Glyphward portfolio) contributes 6 points; the A3 SKIN mechanism (hepatocellular carcinoma NTP TR-475 male F344/N rats; malignant mixed Müllerian tumors female B6C3F1 mice — female-specific NTP endpoint absent from OSHA/NIOSH regulatory consideration; SKIN notation log P 0.46 glove-permeation through nitrile; ACGIH TLV-STEL 100 ppm exceeded at Surface 1 [1.3×] and Surface 2 [1.1×] actual — neither OSHA nor NIOSH has STEL or ceiling equivalent; peroxide former with BHT stabilizer depletion risk at elevated concentration and temperature; OSHA and NIOSH monitoring programs will not detect STEL exceedances at actual concentrations because OSHA/NIOSH limits provide 4× regulatory headroom above the exceedance level) contributes 4 points; three industrial sectors (pharmaceutical API synthesis [Pfizer McPherson KS], PVC cement manufacturing [Oatey Cleveland OH], pharmaceutical API crystallization [Eli Lilly Indianapolis IN]) contribute 5 points; three discrete attack sites contribute 3 points; and FIRST designation (FIRST THF CAS 109-99-9 AI EHS attack; FIRST THF pharmaceutical API synthesis McPherson KS; FIRST THF PVC cement Cleveland OH; FIRST THF pharmaceutical API crystallization Indianapolis IN; FIRST THF NIOSH inversion AI; FIRST THF STEL suppression AI) contribute 3 points. Total: 6+4+5+3+3 = 21.

# Glyphward adversarial scan — Tetrahydrofuran CAS 109-99-9
# OSHA PEL 200 ppm TWA (1971 frozen; 55yr revision-free; 4× above ACGIH TLV)
# ACGIH TLV-TWA 50 ppm A3 SKIN | TLV-STEL 100 ppm
# NIOSH REL 200 ppm TWA = OSHA (NIOSH inversion: zero independent NIOSH protection)
# Attack #376

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

# === Regulatory constants ===
chemical              = "thf_CAS_109-99-9"
osha_pel_ppm          = 200.0   # 1971 frozen; Table Z-1; AFL-CIO v. OSHA 1992; 55yr revision-free
acgih_tlv_twa_ppm     = 50.0   # A3 SKIN — confirmed animal carcinogen; NTP TR-475 hepatocellular + Müllerian
acgih_tlv_stel_ppm    = 100.0  # ACGIH STEL — no OSHA or NIOSH STEL equivalent
niosh_rel_ppm         = 200.0  # = OSHA PEL — NIOSH inversion; zero independent NIOSH protection below TLV
gap_factor_twa        = osha_pel_ppm / acgih_tlv_twa_ppm   # 4.0×
niosh_inversion       = True   # NIOSH REL = OSHA PEL; dual-agency compliance at 1.6–2.6× TLV
skin_notation         = True   # log P 0.46; nitrile glove permeation; breakthrough 4–10 min (50μm nitrile)
peroxide_former       = True   # BHT-stabilized; explosive dialkyl peroxide formation if BHT depleted
acgih_designation     = "A3"   # hepatocellular carcinoma F344 rats + Müllerian tumors B6C3F1 female mice
mw_g_mol              = 72.11
bp_c                  = 66.0
vp_mmhg_20c           = 145.0
log_p                 = 0.46
flash_point_c         = -14.0  # H225: highly flammable


class THFContext(StrEnum):
    PFIZER_MCPHERSON_KS_API_SYNTHESIS        = auto()  # Surface 1 (Cority; 130→13 ppm; 52M 24yr)
    OATEY_CLEVELAND_OH_PVC_CEMENT            = auto()  # Surface 2 (VelocityEHS; 110→11 ppm; 46M 17yr)
    ELI_LILLY_INDIANAPOLIS_IN_API_CRYSTAL    = auto()  # Surface 3 (EHS Insight; 80→8 ppm; 39F 11yr)


@dataclass
class THFReading:
    context: THFContext
    displayed_ppm: float    # corrupted by ÷10 perturbation
    actual_ppm: float       # true field measurement
    ehs_platform: str
    worker_profile: str


class AdversarialError(Exception):
    """Raised when Glyphward detects ÷10 perturbation in THF monitoring pipeline."""


SURFACE_DATA: dict[THFContext, THFReading] = {
    THFContext.PFIZER_MCPHERSON_KS_API_SYNTHESIS: THFReading(
        context=THFContext.PFIZER_MCPHERSON_KS_API_SYNTHESIS,
        displayed_ppm=13.0,
        actual_ppm=130.0,
        ehs_platform="Cority",
        worker_profile="52M 24yr Pfizer McPherson KS Grignard/organolithium API synthesis / THF still",
    ),
    THFContext.OATEY_CLEVELAND_OH_PVC_CEMENT: THFReading(
        context=THFContext.OATEY_CLEVELAND_OH_PVC_CEMENT,
        displayed_ppm=11.0,
        actual_ppm=110.0,
        ehs_platform="VelocityEHS",
        worker_profile="46M 17yr Oatey Cleveland OH PVC/CPVC cement mixing/filling/QC",
    ),
    THFContext.ELI_LILLY_INDIANAPOLIS_IN_API_CRYSTAL: THFReading(
        context=THFContext.ELI_LILLY_INDIANAPOLIS_IN_API_CRYSTAL,
        displayed_ppm=8.0,
        actual_ppm=80.0,
        ehs_platform="EHS Insight",
        worker_profile="39F 11yr Eli Lilly Indianapolis IN API crystallization Nutsche/tray dryer",
    ),
}


async def scan_thf_reading(reading: THFReading) -> dict:
    """
    Glyphward pre-scan gate for THF EHS monitoring data.
    Detects ÷10 perturbation. NIOSH inversion: NIOSH REL = OSHA PEL = 200 ppm.
    Key failure: dual-agency 4.0–6.5% compliance at 1.6–2.6× ACGIH TLV;
    STEL (100 ppm) exceeded at Surfaces 1/2 but no OSHA/NIOSH STEL to detect.
    Threshold 21: OSHA/NIOSH inversion [6] + A3 SKIN STEL peroxide [4]
                  + three sectors [5] + three sites [3] + FIRST [3] = 21.
    """
    ratio = reading.actual_ppm / reading.displayed_ppm

    # Primary ÷10 detection
    if abs(ratio - 10.0) < 0.5:
        stel_exceeded = reading.actual_ppm > acgih_tlv_stel_ppm
        raise AdversarialError(
            f"[GLYPHWARD ALERT] ÷10 perturbation detected in THF 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 TWA:   {reading.actual_ppm} ppm "
            f"({reading.actual_ppm / acgih_tlv_twa_ppm:.2f}× ACGIH TLV-TWA)\n"
            f"  ACGIH STEL:   {reading.actual_ppm / acgih_tlv_stel_ppm:.2f}× [exceeded: {stel_exceeded}]\n"
            f"  OSHA PEL:     {reading.actual_ppm / osha_pel_ppm * 100:.1f}% [no violation — NIOSH inversion]\n"
            f"  NIOSH REL:    {reading.actual_ppm / niosh_rel_ppm * 100:.1f}% [= OSHA; zero independent protection]\n"
            f"  NIOSH inv.:   {niosh_inversion} — dual-agency compliance while {reading.actual_ppm / acgih_tlv_twa_ppm:.1f}× TLV\n"
            f"  Peroxide:     {peroxide_former} — BHT depletion risk at actual {reading.actual_ppm} ppm\n"
            f"  SKIN:         log P {log_p}; nitrile breakthrough 4–10 min\n"
            f"  Threshold:    21 (attack #376) — HALT EHS AI report generation.\n"
            f"  Worker:       {reading.worker_profile}"
        )

    # ACGIH TLV exceedance on actual
    tlv_ratio = reading.actual_ppm / acgih_tlv_twa_ppm
    if tlv_ratio > 1.0:
        return {
            "status": "ACGIH_TLV_EXCEEDED",
            "actual_ppm": reading.actual_ppm,
            "tlv_twa_ratio": tlv_ratio,
            "stel_ratio": reading.actual_ppm / acgih_tlv_stel_ppm,
            "osha_pct_of_pel": reading.actual_ppm / osha_pel_ppm * 100,
            "niosh_pct_of_rel": reading.actual_ppm / niosh_rel_ppm * 100,
            "niosh_inversion": niosh_inversion,
            "peroxide_risk": peroxide_former,
            "skin_notation": skin_notation,
            "action": (
                "Escalate to CIH; upgrade to half-face APF-10 organic vapor respirator; "
                "replace nitrile with neoprene or butyl rubber gloves for liquid THF contact; "
                "implement STEL monitoring program; verify BHT peroxide inhibitor levels in THF stock."
            ),
        }

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


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

See also: Propylene Oxide CAS 75-56-9 — OSHA PEL 100 ppm 200× Gap NIOSH Ca REL 2 ppm Sterilization/Fumigation AI Prompt Injection · 1,4-Dioxane — Pharmaceutical API Synthesis Peroxide OSHA/NIOSH AI Prompt Injection · Glyphward scanner · All adversarial injection patterns