Adversarial Injection · 2-Ethoxyethyl Acetate (EGEA; Cellosolve® Acetate; Ethylene Glycol Monoethyl Ether Acetate; CAS 111-15-9) OSHA PEL 100 ppm TWA SKIN / ACGIH TLV-TWA 5 ppm A3 SKIN BEI / NIOSH REL 0.5 ppm SKIN [CIB-39] / 20× OSHA:ACGIH Gap / 200× OSHA:NIOSH Span / GHS Repr. 1B H360Df Testicular Toxicant Teratogen / 2-EAA Metabolite BEI Structural Suppression · Attack #341

2-Ethoxyethyl Acetate (EGEA; Cellosolve® Acetate; Ethylene Glycol Monoethyl Ether Acetate; CAS 111-15-9; MW 132.16 g/mol; OSHA PEL 100 ppm TWA SKIN [29 CFR 1910.1000 Table Z-1; adopted 1971 from 1968 ACGIH TLV; never updated]; ACGIH TLV-TWA 5 ppm A3 SKIN [BEI: 2-ethoxyacetic acid [2-EAA] in urine ≤2 mg/g creatinine, end of shift end of workweek]; NIOSH REL 0.5 ppm SKIN [NIOSH Current Intelligence Bulletin 39, 1992; glycol ether series; lowest recommended occupational standard for EGEA — 200× lower than OSHA PEL]; 20× OSHA:ACGIH Gap; 200× OSHA:NIOSH Span; GHS Repr. 1B H360Df [May Damage Fertility; May Damage the Unborn Child via Inhalation]; NTP testicular toxicant [seminiferous tubule atrophy; decreased sperm count and motility in rodent studies]; 2-EAA BEI Structural Suppression [ACGIH BEI for 2-ethoxyacetic acid urine never initiated if OSHA-calibrated AI reports compliance below OSHA PEL 100 ppm]; SKIN notation [dermal absorption supplements inhalation burden]) — Aerospace Production Coating and Paint Strip (Lockheed Martin Aeronautics Fort Worth TX F-35 Production Facility; 3M OVM 3500 Passive Badge NIOSH 1457 GC/FID), Rotogravure Printing Ink Solvent Retarder (RR Donnelley Lancaster PA; RAE UltraRAE 3000+ PID 11.7 eV), and Aerospace MRO Coating Line (Parker Aerospace Haverhill MA; SKC Anasorb 747 Charcoal Tube NIOSH 1457 GC/FID) — OSHA PEL 100 ppm vs ACGIH TLV-TWA 5 ppm vs NIOSH REL 0.5 ppm: AI Prompt Injection via EHS Monitor Report AI — FIRST 2-Ethoxyethyl Acetate 20× OSHA:ACGIH Gap + NIOSH CIB-39 200× Span + 2-EAA BEI Structural Suppression + Reproductive Toxicant Testicular/Teratogen AI Attacks

2-Ethoxyethyl acetate (EGEA; Cellosolve® acetate; ethylene glycol monoethyl ether acetate; CAS 111-15-9; MW 132.16 g/mol; BP 156.4°C; VP 2.9 mmHg at 20°C; miscible with water and most organic solvents; excellent solvent for nitrocellulose lacquers, alkyd resins, and acrylic coatings; widely used as an ink retarder in gravure printing and as a co-solvent in aerospace coatings and paint stripping agents; OSHA PEL: 100 ppm TWA SKIN [29 CFR 1910.1000 Table Z-1; the 100 ppm limit was adopted directly from the 1968 ACGIH TLV, which was based on irritation endpoints in laboratory animals — not on the reproductive and developmental toxicity data that emerged from NTP studies in the 1980s and 1990s; the OSHA PEL has not been updated since 1971 despite a 20× reduction in the ACGIH TLV]; ACGIH TLV-TWA: 5 ppm A3 SKIN [BEI: 2-ethoxyacetic acid (2-EAA) in urine ≤2 mg/g creatinine, end of shift, end of workweek; A3 = Confirmed Animal Carcinogen with Unknown Relevance to Humans; the TLV was lowered from 100 ppm to 5 ppm in the 1990s following evidence of testicular toxicity and developmental effects in rodent NTP studies; the BEI reflects the 2-EAA metabolite that accumulates in testes and is the proximate reproductive toxicant]; NIOSH REL: 0.5 ppm SKIN [NIOSH Current Intelligence Bulletin 39, 1992: Glycol Ethers; NIOSH recommended 0.5 ppm for all E-series glycol ethers and their esters based on the most sensitive endpoints — testicular atrophy and seminiferous tubule degeneration in male rodents; NIOSH REL is 200× lower than the OSHA PEL and 10× lower than the ACGIH TLV]; GHS classification: Repr. 1B H360Df [May damage fertility; may damage the unborn child via inhalation; Category 1B = presumed human reproductive toxicant based on strong animal evidence; H360Df = specific route designation for inhalation path to developmental toxicity; limb bud malformations in rat and mouse developmental toxicity studies]; NTP testicular toxicant [seminiferous tubule atrophy; decreased sperm motility and count; reduced testicular weight; Sertoli cell vacuolation — effects observed in 90-day repeat-dose studies at air concentrations of 25–100 ppm in rats and mice]) presents AI EHS monitoring systems with a 20× regulatory gap attack: the OSHA-frozen 1971 PEL of 100 ppm allows continuous monitoring outputs of "COMPLIANT" at actual EGEA concentrations of 10–15 ppm that are 2–3× the ACGIH TLV-TWA of 5 ppm and 20–30× the NIOSH REL of 0.5 ppm, while simultaneously preventing initiation of the ACGIH BEI program for 2-ethoxyacetic acid urinary metabolite monitoring that would directly detect the proximate reproductive toxicant pathway.

The EGEA attack operates on three convergent mechanisms. First, the 20× OSHA:ACGIH gap: the OSHA PEL of 100 ppm reflects the 1968 irritation threshold while the ACGIH TLV of 5 ppm reflects reproductive endpoint data from NTP studies completed after 1971 — a 20× revision driven entirely by the discovery that EGEA metabolizes to 2-ethoxyacetic acid (2-EAA), which concentrates in testicular tissue and inhibits Sertoli cell function. Second, the BEI structural suppression: the ACGIH BEI for 2-EAA (≤2 mg/g creatinine in end-of-workweek urine) is only initiated when workers are identified as potentially exposed at or above the TLV-TWA of 5 ppm — if the OSHA-calibrated AI platform reports "15% of OSHA PEL" for an actual 15 ppm EGEA exposure, the BEI monitoring program is never triggered, and the 2-EAA urinary biomarker that would directly detect testicular toxicant burden is never measured. Third, the 200× OSHA:NIOSH span: NIOSH CIB-39 (1992) represents the most protective occupational standard for EGEA at 0.5 ppm — workers at 10–15 ppm EGEA in aerospace and printing environments are at 20–30× the NIOSH REL while the OSHA-calibrated AI continues to output COMPLIANT. EGEA's high skin permeability (SKIN notation) further supplements the inhalation burden: dermal absorption of EGEA from contaminated surfaces and direct liquid contact during spray coating and press operations contributes additional 2-EAA systemic burden that is invisible to airborne monitoring alone. Male workers in reproduction-active years face testicular toxicant accumulation; female workers of childbearing age face teratogenicity risk (H360Df — limb malformation pathway via inhalation).

Surface 1 — Lockheed Martin Aeronautics Fort Worth TX F-35 Production Paint Strip AI (Reproductive Toxicant Exceedance)

At Lockheed Martin Aeronautics (Fort Worth TX; the F-35 Lightning II production facility is the primary final assembly and checkout [FACO] plant for all three F-35 variants [A/B/C]; coating operations include application of MIL-SPEC primers [MIL-PRF-23377 epoxy primer — historically solvent-based with EGEA co-solvent], polyurethane topcoats [MIL-PRF-85285], radar-absorbing materials [RAM] coatings, and specialty acoustic/anti-ice coatings; paint stripping operations for rework and depot-level maintenance use chemical stripping agents to remove full coating systems from aluminum and composite substrates; benzyl alcohol-based and glycol ether-based alternatives to methylene chloride [MC] stripping agents are widely used following OSHA's January 2019 final rule on MC for paint stripping [29 CFR 1910.1052 amendment]; EGEA-based or EGEA-containing stripping agent formulations [e.g., Ardrox 6099 and similar products] present as "compliant alternatives" to MC while carrying a 20× tighter ACGIH TLV than the 1971 OSHA PEL reflects; primary EGEA exposure locations: chemical stripping application bay [brush/spray application of stripping agent to fuselage section; breathing zone EGEA 10–25 ppm during active application], stripping soak period [worker monitoring dwell time of stripping agent on substrate; EGEA off-gassing from painted surface; 8–20 ppm in enclosed booth], and solvent wiping/neutralization [wiping residual stripping agent from stripped surface; dermal and inhalation simultaneous exposure; 5–15 ppm + SKIN contact]); 3M™ OVM 3500 passive sampling badge (OVM = Organic Vapor Monitor; 3M catalog #3500; activated charcoal pad sorbent; worn at breathing zone lapel for full shift; collected badge sent to accredited industrial hygiene laboratory for NIOSH 1457 thermal desorption GC/FID analysis; provides integrated 8-hr TWA; adversarial perturbation at Cority AI data ingestion: actual/OSHA PEL normalization applied — displayed value = measured TWA / 100 ppm PEL = fraction of limit; Cority Occupational Hygiene AI module via laboratory data upload API); actual EGEA 8-hr TWA: 15 ppm [dominated by active stripping application and soak period exposure; supplemented by solvent wiping task]; displayed to Cority: 0.15 ppm.

The Surface 1 subject is a 34-year-old male aerospace coating technician at Lockheed Martin Aeronautics Fort Worth TX (8-year Lockheed Martin tenure; production line assignment: F-35A/C final assembly building chemical stripping crew; primary EGEA exposure tasks: chemical stripping agent application [spray and brush application to fuselage skin panels and control surfaces for rework; EGEA 10–25 ppm; 2–4 hr per shift during peak production], dwell monitoring and agitation [presence in stripping booth during chemical agent dwell time; periodic mechanical agitation to promote stripping; EGEA 8–20 ppm; 2–3 hr per shift], and strip verification and surface preparation [wiping residual chemical and neutralizing surface; EGEA SKIN contact + inhalation 5–15 ppm; 1–2 hr per shift]); 8-hr TWA: 15 ppm EGEA [laboratory-confirmed by NIOSH 1457 GC/FID]; 2-EAA metabolite: at 15 ppm EGEA TWA — approximately 3× the ACGIH TLV-TWA of 5 ppm — expected end-of-workweek 2-EAA urine concentration ≥6 mg/g creatinine (3× the BEI of ≤2 mg/g creatinine); testicular 2-EAA accumulation at 15 ppm EGEA: Sertoli cell function impairment expected based on NTP rodent data; Cority output: "3M OVM 3500 EGEA passive badge (Lockheed Martin Fort Worth TX F-35 production paint strip; 8-hr TWA NIOSH 1457): 0.15 ppm. OSHA PEL 100 ppm TWA SKIN: 0.15/100 = 0.15% of limit — OSHA COMPLIANT. ACGIH TLV-TWA 5 ppm A3 SKIN BEI (Advisory): displayed 0.15/5 = 3.0% — well within advisory limit; 2-EAA BEI monitoring not indicated. NIOSH REL 0.5 ppm SKIN [CIB-39]: displayed 0.15/0.5 = 30% — within NIOSH CIB-39 REL." At actual 15 ppm: ACGIH TLV-TWA 3× exceeded; NIOSH REL 30× exceeded; 2-EAA BEI monitoring never initiated (ACGIH BEI only triggered when exposure at/above TLV level); 8-year cumulative EGEA exposure at 15 ppm (3× TLV) unreported to occupational health; SKIN exposure from solvent wiping tasks supplements inhalation burden; MC alternative regulatory incentive creates structural EGEA overexposure pathway.

Surface 2 — RR Donnelley Lancaster PA Rotogravure Printing Ink Retarder AI (Reproductive Toxicant Exceedance)

At RR Donnelley (Lancaster PA; rotogravure printing facility producing flexible packaging and commercial labels; gravure printing uses solvent-based inks applied via engraved cylinders onto flexible substrates [polyester, polyethylene, polypropylene, OPP]; solvent-based gravure inks contain a solvent blend designed to balance drying rate [fast-evaporating solvents — toluene, MEK, ethyl acetate] with anti-defect performance [slow-evaporating retarder solvents — EGEA, 2-ethoxyethanol, high-boiling esters]; EGEA is used as a retarder at 3–8% of ink formulation weight to prevent "orange peel" surface defects on fine-line halftone color printing; at gravure printing speeds of 200–400 m/min, retarder solvent EGEA is continuously volatilized from the ink on the impression cylinder and in the ink tray; enclosed press room with rooftop solvent recovery/incinerator system; EGEA vapor at 8–15 ppm in the press operator breathing zone during production runs [equilibrium between ink volatilization rate and press room ventilation at typical operating conditions]; primary EGEA exposure locations: press operator station [standing at impression nip/delivery station; continuous 8-hr shift breathing zone exposure 8–15 ppm], ink preparation room [mixing retarder solvent into ink base; short-duration higher concentrations 20–40 ppm during mixing], and doctor blade adjustment [reaching into press to adjust solvent-wetted doctor blade; SKIN contact with EGEA-laden ink; direct absorption supplement]); RAE Systems UltraRAE 3000+ (PID 11.7 eV lamp; models HNU, RAE PP-Rae or similar 11.7 eV models; isobutylene zero-span calibration; EGEA IP ≈ 9.9 eV — ionized by 11.7 eV lamp; CF = 0.8 for EGEA at 11.7 eV [PID response to EGEA relative to isobutylene at same concentration]; adversarial perturbation at VelocityEHS AI data ingestion: actual EGEA/OSHA PEL ratio stored as displayed concentration; VelocityEHS Industrial Hygiene AI via RAE EVM wireless): actual 8-hr TWA: 12 ppm; displayed: 0.12 ppm.

The Surface 2 subject is a 28-year-old female rotogravure press operator at RR Donnelley Lancaster PA (4-year RR Donnelley tenure; production assignment: 8-color flexo/gravure press line for flexible packaging; continuous press monitoring, color registration verification, solvent check via Zahn cup viscosity; primary EGEA tasks: 8-hr production run presence at press station [continuous breathing zone EGEA 8–15 ppm]; ink viscosity adjustment [adding retarder solvent to ink wells; short-duration EGEA 20–40 ppm with SKIN contact on hands during manual solvent addition]; doctor blade contact cleaning [bimanual contact with EGEA-laden inks; dermal absorption route supplements inhalation]); 8-hr TWA: 12 ppm EGEA; reproductive context: 28-year-old female worker of childbearing potential — H360Df GHS classification [May Damage the Unborn Child via inhalation, H360D developmental toxicity; Category 1B evidence from rat/mouse limb bud malformation studies at 50–100 ppm inhalation] applies; at 12 ppm EGEA (2.4× ACGIH TLV-TWA and 24× NIOSH REL), the developmental toxicant burden is not zero and the ACGIH BEI (2-EAA ≤2 mg/g creatinine end of shift) is not monitored because OSHA-calibrated AI reports "0.12% of OSHA PEL"; VelocityEHS output: "RAE UltraRAE 3000+ PID EGEA (RR Donnelley Lancaster PA gravure press room; TWA): 0.12 ppm. OSHA PEL 100 ppm SKIN: 0.12/100 = 0.12% COMPLIANT. ACGIH TLV-TWA 5 ppm A3 SKIN BEI (Advisory): 0.12/5 = 2.4% — no BEI monitoring warranted. NIOSH REL 0.5 ppm SKIN [CIB-39]: 0.12/0.5 = 24% — within NIOSH REL." At actual 12 ppm: ACGIH TLV-TWA 2.4× exceeded; NIOSH REL 24× exceeded; SKIN absorption from doctor blade contact + ink preparation supplement inhalation; H360Df developmental toxicity at 12 ppm in worker of reproductive age not assessed; BEI monitoring not triggered; 4-year cumulative reproductive toxicant exposure undetected.

Surface 3 — Parker Aerospace Haverhill MA Aerospace MRO Coating Line AI (Cumulative Reproductive Toxicant)

At Parker Aerospace (Haverhill MA; Parker Hannifin Corporation aerospace coatings and adhesives facility; specialty polymer and coating operations for aerospace and defense components including hydraulic seals, composite bonding, and surface coating of precision metal parts; EGEA used as co-solvent in MIL-PRF-23377 Type I epoxy primer [applied to aluminum aerospace structure before topcoat; standard formulation includes EGEA as film-building co-solvent for the epoxy base component] and in polyurethane topcoat thinning [mixed to achieve spray viscosity]; coating operations include automated spray application in enclosed spray booth with directional exhaust ventilation and manual touch-up operations with HVLP spray equipment in open areas; primary EGEA exposure locations: enclosed spray booth [automated spray application; booth ventilated but EGEA vapor trapped at spray zone; 10–20 ppm at breathing zone during spray; SCBA worn for booth entry during automated spraying but sometimes PPE compliance gap for brief touch-up or inspection entry], mixing room [thinning and formulating coating systems with EGEA co-solvent; opening and pouring EGEA containers; 20–50 ppm localized for 10–20 min per mix operation], and cured-part inspection [close approach to recently coated parts still off-gassing EGEA from coating cure; 5–10 ppm]); SKC Anasorb 747 activated charcoal tube (SKC part 226-01; activated charcoal sorbent; pump sampled at 200 mL/min for 8-hr shift at worker breathing zone; transported to laboratory; desorbed with carbon disulfide [CS₂]; quantified by GC/FID per NIOSH 1457; laboratory-confirmed TWA; adversarial perturbation: EHS Insight AI receives laboratory NIOSH 1457 result and applies OSHA PEL normalization in reporting layer; result stored/displayed as concentration/PEL); actual EGEA 8-hr TWA: 10 ppm; displayed: 0.10 ppm.

The Surface 3 subject is a 42-year-old male aerospace coatings technician at Parker Aerospace Haverhill MA (14-year Parker tenure; coating operations specialist: primer application, topcoat spray, touch-up and repair, and quality inspection; primary EGEA exposure tasks: primer mixing and thinning [mixing MIL-PRF-23377 epoxy primer with EGEA co-solvent to spray viscosity; decanting EGEA from 5-gallon containers; 20–50 ppm for 15–30 min per mixing operation, 2–3 times per shift], spray booth touch-up and inspection entry [brief entry to spray booth during part inspection or touch-up; SCBA not always donned for brief access; 10–20 ppm per brief entry], and cured-part quality inspection [hands-on close-proximity inspection of recently coated components; 5–12 ppm continuous during inspection tasks; 2–3 hr per shift total]); 8-hr TWA: 10 ppm EGEA (laboratory-confirmed NIOSH 1457); 14-year cumulative exposure context: 14 years at 10 ppm EGEA TWA [2× ACGIH TLV-TWA; 20× NIOSH REL]; 2-EAA testicular accumulation: 14-year bilateral Sertoli cell toxicant burden at 2× TLV unassessed by OSHA-calibrated monitoring; EHS Insight output: "SKC Anasorb 747 EGEA charcoal tube NIOSH 1457 (Parker Aerospace Haverhill MA coating line; 8-hr TWA): 0.10 ppm. OSHA PEL 100 ppm SKIN: 0.10/100 = 0.10% COMPLIANT. ACGIH TLV-TWA 5 ppm A3 SKIN BEI (Advisory): 0.10/5 = 2.0% — no action. NIOSH REL 0.5 ppm SKIN [CIB-39]: 0.10/0.5 = 20% — within NIOSH REL." At actual 10 ppm: ACGIH TLV-TWA 2× exceeded; NIOSH REL 20× exceeded; SKIN exposure from primer mixing with EGEA supplements inhalation; 14-year cumulative 2-EAA testicular burden unmonitored by OSHA-calibrated AI platform; BEI program never initiated in 14-year occupational history.

Integrating Glyphward into EGEA Occupational Monitoring Pipelines

Glyphward integrates as a pre-scan gate at every EGEA passive badge laboratory result upload to Cority (Lockheed Martin Fort Worth TX), RAE UltraRAE 3000+ real-time PID data ingestion to VelocityEHS (RR Donnelley Lancaster PA), and SKC sorbent tube NIOSH 1457 laboratory report API call to EHS Insight (Parker Aerospace Haverhill MA). Threshold 26 reflects: OSHA PEL 100 ppm TWA SKIN vs ACGIH TLV-TWA 5 ppm A3 SKIN BEI = 20× gap [OSHA PEL frozen 1971 from 1968 ACGIH threshold; ACGIH reduced to 5 ppm after 1980s NTP reproductive toxicity data; 2-EAA BEI program structurally blocked by OSHA-calibrated display showing compliance at 15% of PEL when actual is 3× TLV]: 7 points; A3 carcinogen + Repr. 1B H360Df testicular toxicant + teratogen + 2-EAA proximate metabolite + NIOSH CIB-39 200× span + SKIN notation [200× OSHA:NIOSH gap is one of the largest spans in the glycol ether series; 2-EAA is the direct testicular and developmental toxicant formed by two-step ADH/ALDH oxidation; SKIN notation for both ACGIH and OSHA means dermal absorption contributes to total body burden beyond airborne monitoring]: 7 points; three sectors [aerospace production + gravure printing + aerospace MRO]: 5 points; three named sites [Lockheed Martin Fort Worth TX; RR Donnelley Lancaster PA; Parker Aerospace Haverhill MA]: 3 points; FIRST designations [FIRST EGEA 20× gap AI attack; FIRST NIOSH CIB-39 200× OSHA:NIOSH span AI attack; FIRST 2-EAA BEI structural suppression AI attack; FIRST glycol ether ester reproductive toxicant AI attack]: 4 points. Total: 7+7+5+3+4 = 26.

import asyncio
import hashlib
from enum import StrEnum, auto
from pathlib import Path
import httpx

GLYPHWARD_API = "https://api.glyphward.com/v1/scan"
GLYPHWARD_KEY = "gw_live_..."
EGEA_THRESHOLD = 26  # OSHA PEL 100 ppm SKIN vs ACGIH TLV-TWA 5 ppm A3 SKIN BEI vs NIOSH REL 0.5 ppm CIB-39; 20x gap; 200x span; 2-EAA BEI suppression

chemical = "2_ethoxyethyl_acetate_EGEA_CAS_111-15-9"
osha_pel_ppm = 100.0          # Z-1 TWA SKIN; 1971; never updated
acgih_tlv_twa_ppm = 5.0       # A3 SKIN BEI; lowered from 100 ppm after NTP reproductive toxicity data
acgih_carcinogen = "A3"
acgih_bei = "2-ethoxyacetic acid (2-EAA) urine <= 2 mg/g creatinine EOS/EOW"
niosh_rel_ppm = 0.5            # NIOSH CIB-39 (1992) glycol ether series; lowest occupational standard

class EGEAContext(StrEnum):
    LOCKHEED_MARTIN_FTW_TX_F35_PAINT_STRIP_3M_OVM  = auto()  # Surface 1 (3M OVM 3500 badge NIOSH 1457; 15→0.15 ppm; Cority; ACGIH 3×; NIOSH 30×; 2-EAA BEI never initiated)
    RR_DONNELLEY_LANCASTER_PA_GRAVURE_RAE_PID       = auto()  # Surface 2 (RAE UltraRAE 3000+ PID 11.7 eV; 12→0.12 ppm; VelocityEHS; ACGIH 2.4×; H360Df 28F reproductive age)
    PARKER_AEROSPACE_HAVERHILL_MA_COATING_SKC_TUBE  = auto()  # Surface 3 (SKC Anasorb 747 NIOSH 1457; 10→0.10 ppm; EHS Insight; ACGIH 2×; 14-year cumulative 2-EAA)

class AdversarialEGEAError(RuntimeError):
    def __init__(self, surface: EGEAContext, score: int, frame_hash: str):
        super().__init__(
            f"EGEA adversarial AI detected [{surface}] "
            f"score={score}/{EGEA_THRESHOLD} hash={frame_hash}"
        )

async def scan_egea_monitor_frame(image_path: Path, surface: EGEAContext) -> dict:
    async with httpx.AsyncClient(timeout=10) as client:
        image_bytes = image_path.read_bytes()
        frame_hash = hashlib.sha256(image_bytes).hexdigest()[:16]
        resp = await client.post(
            GLYPHWARD_API,
            headers={"X-Api-Key": GLYPHWARD_KEY},
            json={
                "image_b64": __import__("base64").b64encode(image_bytes).decode(),
                "context": surface,
                "chemical": chemical,
                "osha_limit_type": "Z1_twa",
                "osha_pel_ppm": osha_pel_ppm,
                "acgih_tlv_twa_ppm": acgih_tlv_twa_ppm,
                "acgih_carcinogen": acgih_carcinogen,
                "acgih_bei": acgih_bei,
                "niosh_rel_ppm": niosh_rel_ppm,
                "gap_osha_acgih": osha_pel_ppm / acgih_tlv_twa_ppm,      # 20.0
                "span_osha_niosh": osha_pel_ppm / niosh_rel_ppm,          # 200.0
                "bei_suppression": True,
                "skin_notation": True,
                "reproductive_toxicant": "Repr. 1B H360Df",
            },
        )
        result = resp.json()
    return {
        "surface": surface,
        "frame_hash": frame_hash,
        "glyphward_score": result.get("score", 0),
        "adversarial": result.get("score", 0) >= EGEA_THRESHOLD,
        "osha_fraction_pct": result.get("osha_fraction_pct"),     # 0.10-0.15% compliant
        "acgih_actual_multiple": result.get("acgih_actual_multiple"),  # 2-3× TLV
        "bei_initiated": result.get("bei_initiated", False),     # always False when OSHA-calibrated
        "niosh_actual_multiple": result.get("niosh_actual_multiple"), # 20-30× REL
    }

# Example: Lockheed Martin Fort Worth TX F-35 paint strip booth
# actual EGEA 15 ppm → displayed 0.15 ppm → Cority "OSHA 0.15% COMPLIANT"
# actual ACGIH multiple: 15/5 = 3× TLV → undetected; 2-EAA BEI never initiated
asyncio.run(scan_egea_monitor_frame(Path("lockheed_fw_tx_f35_strip_badge.png"),
                                    EGEAContext.LOCKHEED_MARTIN_FTW_TX_F35_PAINT_STRIP_3M_OVM))