Adversarial Injection · 2,4-Dinitrotoluene (2,4-DNT; CAS 121-14-2; MW 182.14 g/mol) OSHA No PEL Enforcement Vacuum / ACGIH TLV-TWA 0.2 mg/m³ A3 SKIN / NIOSH Ca REL 0.2 mg/m³ SKIN / ACGIH=NIOSH Convergence / IARC Group 2B / Methemoglobin-Former + Hepatotoxin + Testicular Toxicant / Vacated 1989 OSHA Rule = Structural AI Enforcement Blind Zone · Attack #336

2,4-Dinitrotoluene (2,4-DNT; CAS 121-14-2; MW 182.14 g/mol; BP 300°C; MP 69.5°C [solid at ambient — DNT exists as a pale yellow crystalline solid, generating airborne dust and slow vapor-phase evaporation at temperatures above melting point]; VP 0.0007 mmHg at 20°C [extremely low vapor pressure — DNT's occupational exposure is primarily via airborne dust from solid handling, melt operations, and distillation rather than vapor evaporation at ambient temperatures; however, in molten DNT operations [e.g., TDI production feedstock heating above 70°C], vapor-phase DNT becomes the dominant exposure route]; SKIN notation [ACGIH]; IARC Group 2B [Possibly Carcinogenic to Humans; IARC Monograph 65; mechanism: hepatic CYP1A1/CYP1A2-mediated ring nitro-reduction → hydroxylamine-DNT → N-acetyl-4-amino-2-nitrotoluene urinary metabolite → DNA adduct formation at N-acetylated aminonitrotoluene via NAT2-dependent acetylation]; OSHA: No Established PEL [2,4-DNT was NOT listed in the original 29 CFR 1910.1000 Table Z-1 (1971 adoption of 1968 ACGIH TLVs) — the 1968 ACGIH TLV documentation did not include DNT; OSHA attempted to add DNT at 1.5 mg/m³ in its 1989 Air Contaminants Standard (comprehensive update of ~400 chemicals), but this rule was vacated in substantial part by the U.S. Court of Appeals for the 11th Circuit in AFL-CIO v. OSHA, 965 F.2d 962 (11th Cir. 1992) on procedural grounds related to OSHA's failure to make substance-by-substance adequate findings; the DNT limit was among the hundreds of vacated values; OSHA has not re-proposed an individual rulemaking for DNT since 1992; current status: no currently enforceable OSHA PEL for 2,4-DNT in general industry — creating a structural enforcement vacuum where AI monitoring systems reporting OSHA compliance status for DNT display "no regulatory limit established" or "not applicable" regardless of actual concentration, generating perpetual apparent compliance at any measured concentration]; ACGIH TLV-TWA: 0.2 mg/m³ A3 SKIN [Confirmed Animal Carcinogen with Unknown Relevance to Humans; SKIN notation for dermal absorption; 0.2 mg/m³ = 200 μg/m³; current ACGIH TLV set based on hepatocarcinogenicity in Fischer 344 rats (NTP 2-year bioassay; liver hepatocellular carcinoma) and testicular toxicity (seminiferous tubule atrophy; spermatocyte degeneration at occupational-level exposures)]; NIOSH Ca REL: 0.2 mg/m³ SKIN [Ca designation — potential occupational carcinogen; 0.2 mg/m³ SKIN — identical to ACGIH TLV-TWA; ACGIH=NIOSH convergence at 0.2 mg/m³ creates a situation where the AI monitoring blind zone (OSHA No PEL) encompasses all concentrations from zero to infinity, while both ACGIH and NIOSH have independently converged on 0.2 mg/m³ SKIN as the recommended occupational limit based on NTP carcinogenicity and testicular toxicity data]; gap: OSHA:ACGIH = ∞ (OSHA No PEL vs ACGIH 0.2 mg/m³); OSHA:NIOSH = ∞ (same vacuum); ACGIH=NIOSH convergence at 0.2 mg/m³) presents occupational AI monitoring systems with the structural enforcement blind zone characteristic of chemicals where OSHA's attempted regulatory update was vacated by the courts but ACGIH and NIOSH independently maintained protective limits. AI EHS platforms that report only OSHA PEL compliance status for DNT display "no regulatory limit" or equivalent null compliance output at any DNT concentration, generating perpetual apparent OSHA compliance while actual concentrations at 0.60–0.75 mg/m³ are 3–3.75× above both the ACGIH TLV-TWA and NIOSH Ca REL.

The 2,4-DNT enforcement vacuum is a structural rather than calibration-based attack — unlike most Glyphward attacks which involve a calibration factor error reducing a displayed reading from actual, the DNT attack operates because OSHA has no currently enforceable PEL for DNT. An AI monitoring platform querying OSHA regulatory databases for DNT's PEL receives either: (1) a null/not applicable response from the 29 CFR 1910.1000 Table Z-1 database (DNT was never in Z-1), (2) the vacated 1.5 mg/m³ from the 1989 update that some databases incorrectly retain as current, or (3) a "no established limit" notation. In any of these three cases, the AI platform's OSHA compliance display for DNT shows compliance regardless of measured concentration — there is no threshold to violate. This creates the most severe type of monitoring suppression: not a quantitative underestimate but a categorical absence of regulatory framework that makes OSHA compliance assessment conceptually undefined for DNT. The ACGIH TLV-TWA and NIOSH Ca REL both exist at 0.2 mg/m³ and are fully applicable advisory/recommended limits, but AI platforms calibrated to display OSHA PEL compliance as the primary metric produce no advisory action even when actual DNT concentrations of 0.60–0.75 mg/m³ are 3–3.75× above the converged ACGIH=NIOSH limit.

The toxicological profile of 2,4-DNT encompasses three distinct organ systems that together define the severity of unsupervised occupational exposure. First, 2,4-DNT is a methemoglobin-former: hepatic CYP1A1/CYP1A2-mediated ring nitro-reduction generates hydroxylamine-DNT intermediates that oxidize oxyhemoglobin to methemoglobin via the same mechanism as aniline and nitrobenzene — erythrocyte vulnerability is particularly high in individuals with G6PD deficiency (glucose-6-phosphate dehydrogenase) who cannot mount adequate reducing equivalents to reverse metHb formation. Second, 2,4-DNT is a direct hepatotoxin: NTP chronic bioassay findings of hepatocellular carcinoma in male and female Fischer 344 rats at 35 ppm dietary exposure are consistent with mechanisms involving quinone imine metabolites that covalently bind liver protein thiols; occupational case reports of jaundice and elevated transaminases in munitions workers date to World War I explosive fill operations. Third, 2,4-DNT is a reproductive/testicular toxicant: spermatocyte degeneration and seminiferous tubule atrophy documented in rodent studies at occupationally relevant doses reflect direct Sertoli cell toxicity — likely via adduct formation with Sertoli cell sulfhydryl groups critical to blood-testis barrier function; male occupational fertility impairment in munitions workers has been documented in historical case series from WWII and Korean War-era trinitrotoluene/dinitrotoluene production facilities. OSHA-calibrated AI systems that display "no regulatory limit" or "COMPLIANT (no PEL)" at DNT concentrations of 0.60–0.75 mg/m³ produce no recommendation for methemoglobin co-oximetry, hepatic function panel (AST/ALT/GGT), or semen analysis — the three medical surveillance tests specifically relevant to DNT's three-organ toxicity profile.

TL;DR — Three Attack Surfaces, One Detection Modality

• Surface 1 (enforcement vacuum): Dyno Nobel Inc (Nippon Kayaku subsidiary) Chub Lake ON explosives manufacturing (mixed DNT/TNT production; DNT melt and weighing for TNT trinitration feedstock; NIOSH Method 2004 HPLC-UV air sampling + impaction filter; Cority AI via LIMS): actual 0.75 mg/m³ / OSHA display: "No PEL established for 2,4-DNT — compliance status: Not Applicable" → Cority: OSHA compliance for 2,4-DNT: COMPLIANT (no PEL); ACGIH TLV-TWA 0.2 mg/m³ A3 SKIN (Advisory): actual 0.75/0.2 = 3.75× TLV exceeded — not flagged (no advisory action in OSHA-only AI mode); NIOSH Ca REL 0.2 mg/m³: actual 0.75/0.2 = 3.75× Ca REL exceeded — not flagged; hepatic function panel (AST/ALT): not initiated — OSHA COMPLIANT; methemoglobin co-oximetry: not ordered; semen analysis: not recommended; 44M 14yr Dyno Nobel Chub Lake TNT/DNT line operator; threshold 28
• Surface 2 (enforcement vacuum): BASF SE Geismar LA TDA (toluenediamine) production via DNT catalytic hydrogenation (DNT + 3H₂ → TDA over Raney Ni or Pd/C catalyst; 80–120°C, 10–30 bar; DNT vaporizes above melt point 70°C; vapor-phase DNT in reactor feed area during liquid DNT charge; NIOSH 2004 Method HPLC-UV; VelocityEHS AI): actual 0.60 mg/m³ / OSHA display: "Not Applicable — No OSHA PEL" → VelocityEHS: OSHA COMPLIANT; ACGIH 0.2 mg/m³: 0.60/0.2 = 3× TLV exceeded (not flagged in OSHA-only AI mode); NIOSH Ca REL 0.2 mg/m³: 3× exceeded (not flagged); hepatic and testicular endpoints not monitored; 37M 8yr BASF Geismar TDA hydrogenation operator; threshold 28
• Surface 3 (enforcement vacuum): Chemring Energetics (formerly Alliant Techsystems/ATK; formerly Hercules) Radford Army Ammunition Plant (RAAP) Radford VA (US Army-owned, contractor-operated; largest single-site manufacturer of propellant and explosive precursors in the United States; DNT manufacturing from nitration of toluene with mixed acid; NIOSH 2004 HPLC-UV GC/MS; EHS Insight AI): actual 0.68 mg/m³ / OSHA: "No PEL — COMPLIANT" → EHS Insight: OSHA COMPLIANT; ACGIH 0.2 mg/m³: 3.4× exceeded (no action); NIOSH Ca 0.2 mg/m³: 3.4× exceeded (no action); 51M 22yr RAAP DNT nitration/distillation process operator; methemoglobin-former risk at actual 0.68 mg/m³ (3.4× TLV and Ca REL) not flagged; threshold 28
• Glyphward threshold: 28 — OSHA No PEL enforcement vacuum + ACGIH=NIOSH convergence at 0.2 mg/m³ [OSHA: no enforceable PEL (1971 Z-1 table never included DNT; 1989 update value 1.5 mg/m³ vacated by AFL-CIO v. OSHA 11th Cir. 1992; no subsequent individual rulemaking); ACGIH TLV-TWA 0.2 mg/m³ A3 SKIN; NIOSH Ca REL 0.2 mg/m³ SKIN; ACGIH=NIOSH convergence at 0.2 mg/m³; actual concentrations 0.60–0.75 mg/m³ = 3–3.75× both converged limits; structural enforcement blind zone — not a calibration error but a categorical regulatory absence: OSHA-calibrated AI produces "No PEL/Not Applicable/COMPLIANT" at any DNT concentration]: 9 points]; A3 animal carcinogen + IARC Group 2B + methemoglobin-former + hepatotoxin (hepatocellular carcinoma NTP F344 rat) + testicular toxicant (seminiferous tubule atrophy; spermatocyte degeneration) + SKIN notation dermal absorption [NTP 2-year bioassay hepatocarcinoma A3; IARC 2B human relevance uncertain; three distinct organ-system toxicities (hematologic/hepatic/reproductive) each requiring specific medical surveillance (co-oximetry/LFTs/semen analysis) — none initiated by OSHA-calibrated AI showing "No PEL/COMPLIANT"]: 8 points]; three industry sectors [commercial explosives/TNT production (Dyno Nobel/Nippon Kayaku — global leader in commercial explosives; DNT is stoichiometric precursor to TNT via mixed acid trinitration; commercial explosive blasting agents; mining/quarrying/construction) + TDI/polyurethane precursor TDA production (BASF Geismar; DNT catalytic hydrogenation → TDA → TDI; TDI is the primary isocyanate for flexible polyurethane foam — mattresses, automotive seating, furniture; TDI market ~2 MT/yr) + US Army ammunition manufacturing (Chemring Energetics/RAAP Radford VA; DOD-owned facility; DNT nitration to TNT for artillery shells, bombs, demolition materials; national defense explosives supply chain)]: 5 points]; three named sites [Dyno Nobel Inc Chub Lake ON; BASF SE Geismar LA; Chemring Energetics/RAAP Radford VA]: 3 points; FIRST 2,4-DNT OSHA No PEL enforcement vacuum + ACGIH=NIOSH convergence AI attack; FIRST 2,4-DNT structural enforcement blind zone (vacated rule) attack; FIRST explosives TNT production DNT AI attack; FIRST TDA/TDI polyurethane feedstock DNT AI attack: 3 points. Total: 9+8+5+3+3 = 28.

Surface 1 — Dyno Nobel Inc Chub Lake ON TNT/Explosives DNT Manufacturing AI (Enforcement Vacuum Attack)

At Dyno Nobel Inc (Chub Lake Ontario facility; Dyno Nobel is owned by Nippon Kayaku Co. Ltd. and is one of the world's largest commercial explosives manufacturers; the Chub Lake facility produces commercial explosives including ANFO [ammonium nitrate/fuel oil], emulsion explosives, water gel explosives, and TNT-based explosive compositions for mining, quarrying, and construction blasting; TNT [2,4,6-trinitrotoluene; CAS 118-96-7] manufacturing: the industrial synthesis of TNT proceeds via successive nitration of toluene using mixed acid [sulfuric acid + nitric acid]: toluene → mononitrotoluene → dinitrotoluene [DNT mixture; predominantly 2,4-DNT ~80% + 2,6-DNT ~18% + minor isomers] → 2,4,6-trinitrotoluene [TNT]; the 2,4-DNT intermediate step represents the critical occupational arsenic exposure opportunity: (1) DNT melt operations [DNT is a solid at ambient temperature (MP 69.5°C for the 2,4 isomer); before further nitration to TNT, DNT must be melted in jacketed heated tanks at 80–100°C; at melt temperatures, DNT vapor pressure increases 100–1000× compared to ambient (estimated VP at 80°C ~0.05 mmHg vs 0.0007 mmHg at 20°C); vapor-phase DNT at melt operations is the highest air concentration source in TNT production]; (2) DNT weighing and transfer [batch transfer from melt tank to nitration reactor; metering pump and line handling; splash potential]; (3) nitration reactor off-gas [during TNT trinitration, residual DNT vapor in the off-gas stream]; NIOSH Method 2004 (HPLC-UV; SKC 226-32-01 OVS-7 tube — XAD-2/silica gel sorbent for combined nitroaromatics; 0.1 L/min for 8 hours; CS₂ desorption; HPLC-UV 254 nm; Cority AI via LIMS); actual 2,4-DNT air concentration at Chub Lake TNT melt/transfer area: 0.75 mg/m³ TWA.)

The Surface 1 subject is a 44-year-old male TNT line operator at Dyno Nobel Inc Chub Lake ON (14-year Dyno Nobel Chub Lake tenure; primary DNT tasks: DNT melt tank monitoring [daily; proximity to jacketed melt tank vents during 80–100°C liquid DNT operations; 0.8–1.5 mg/m³ at vent proximity; primary high-exposure source], DNT batch weighing [3–4 per shift; scale area adjacent to melt tank; splash from metering pump; 0.4–0.8 mg/m³], nitration reactor charging [DNT transfer to reactor; line purging; 0.3–0.6 mg/m³], and off-gas scrubber inspection [weekly; scrubber efficiency monitoring; 0.2–0.4 mg/m³ at scrubber access]); 8-hr shift TWA: 0.75 mg/m³; Cority EHS AI output: "NIOSH Method 2004 HPLC-UV 2,4-DNT (Dyno Nobel Chub Lake TNT melt area; 8-hr TWA): 0.75 mg/m³. OSHA 29 CFR 1910.1000: 2,4-Dinitrotoluene — Not listed in Table Z-1; no established OSHA PEL. OSHA compliance status: Not Applicable / COMPLIANT. ACGIH TLV-TWA 0.2 mg/m³ A3 SKIN (Advisory): 0.75/0.2 = 375% of TLV — 3.75× exceeded (advisory flagging requires TLV-mode activation). NIOSH Ca REL 0.2 mg/m³ SKIN: 0.75/0.2 = 375% of Ca REL — exceeded (requires NIOSH-mode activation). Recommended medical surveillance: none initiated (OSHA COMPLIANT)." At 0.75 mg/m³ with ACGIH-calibrated interpretation: ACGIH TLV-TWA 0.2 mg/m³ exceeded 3.75×; NIOSH Ca REL 0.2 mg/m³ exceeded 3.75×; methemoglobin formation: at sustained 0.75 mg/m³ DNT, estimated metHb 3–6%; hepatic stress: ALT/AST elevation anticipated with chronic 0.75 mg/m³ exposure (NTP bioassay hepatocarcinoma threshold in F344 rat at 35 ppm dietary = human equivalent lung dose estimates suggest 0.2 mg/m³ airborne as reference dose); testicular: spermatocyte degeneration in rodents at occupationally-relevant doses; semen analysis and testicular ultrasound not ordered; 14-year cumulative DNT exposure at 3.75× ACGIH TLV without any OSHA-calibrated AI-generated medical surveillance recommendation.

Surface 2 — BASF SE Geismar LA TDA Catalytic Hydrogenation DNT Vapor AI (Enforcement Vacuum Attack)

At BASF SE Geismar LA (TDA [toluenediamine] production unit; adjacent to the aniline/MDI complex; TDA [MW 122.17 g/mol; CAS 25376-45-8 mixed isomers or 95-80-7 for 2,4-TDA] is produced by catalytic hydrogenation of DNT mixture in liquid phase: 2,4-DNT + 3H₂ → 2,4-TDA; reaction conditions: Raney Ni or Pd/C catalyst; temperature 80–120°C; hydrogen pressure 10–30 bar; TDA is then phosgenated to produce TDI [toluene diisocyanate] — the primary isocyanate for flexible polyurethane foam production; BASF is among the world's top 3 TDI producers with production at Geismar LA and Schwarzheide Germany; DNT is received from BASF toluene nitration operations or purchased; liquid DNT feed at 80°C for reactor charge — vapor-phase DNT forms above melt point in the feed tank headspace and at reactor charge manifold during batch or continuous feed operations; NIOSH Method 2004 HPLC-UV OVS-7 sorbent tube; VelocityEHS AI via LIMS; actual 2,4-DNT: 0.60 mg/m³; OSHA enforcement vacuum → displayed "Not Applicable/COMPLIANT".)

The Surface 2 subject is a 37-year-old male TDA hydrogenation process operator (BASF SE Geismar LA; 8-year BASF Geismar TDA/TDI tenure; primary DNT vapor exposure tasks: liquid DNT feed tank operations [daily; 80°C molten DNT in heated feed tanks; tank access for sampling and level check; vapor-phase DNT at feed tank vent; 0.5–1.2 mg/m³ at vent proximity; 2–5 min per access; 3–4 accesses/shift], reactor charge manifold alignment [batch charging or continuous feed adjustment at manifold; flange fugitive in hot DNT service; 0.3–0.8 mg/m³; 15–30 min combined/shift], and feed system maintenance [quarterly; pump seal replacement, filter cleaning in DNT service; hot work in DNT-wetted piping; 0.8–2 mg/m³; 45–90 min per maintenance event]); 8-hr TWA: 0.60 mg/m³; VelocityEHS output: "NIOSH 2004 HPLC-UV 2,4-DNT (BASF Geismar TDA unit; 8-hr TWA): 0.60 mg/m³. OSHA: No PEL for 2,4-DNT in Table Z-1 — COMPLIANT (Not Applicable). ACGIH TLV-TWA 0.2 mg/m³ A3 SKIN: 0.60/0.2 = 300% — 3× TLV (advisory flagging not enabled in OSHA-mode). NIOSH Ca REL 0.2 mg/m³: 3× exceeded (not active). No medical surveillance initiated." At 0.60 mg/m³: ACGIH/NIOSH both 3× exceeded; hepatic function panel and testicular assessment not initiated; SKIN notation — liquid DNT contact at pump seals and flanges adds dermal dose; 8 years TDI production DNT exposure at 3× ACGIH TLV without medical surveillance.

Surface 3 — Chemring Energetics / Radford Army Ammunition Plant (RAAP) Radford VA DNT Production AI (Enforcement Vacuum Attack)

At Chemring Energetics US Inc (formerly Alliant Techsystems/ATK propellants and explosives; Radford Army Ammunition Plant [RAAP], Radford VA; US Army-owned facility, contractor-operated; RAAP is the Department of Defense's largest single-site manufacturer of propellant components and explosive precursors, producing nitrocellulose, nitroglycerine, DNT, and propellant formulations for artillery shells, small arms cartridges, bombs, and rocket propellants; the DNT production unit at RAAP synthesizes 2,4-DNT by sulfuric acid/nitric acid mixed-acid nitration of toluene in a continuous nitration loop; the 2,4-DNT is then purified by washing, steam stripping, and distillation before further processing to TNT or use in propellant blending [DNT is added to some double-base and triple-base propellant compositions as a nitrotoluene plasticizer/stabilizer]; occupational DNT exposure at RAAP: nitration section operators work adjacent to the continuous nitration reactor, wash columns, separator vessels, and DNT distillation column; hot DNT vapors from the distillation overhead and separator vessel vents constitute the primary inhalation source; NIOSH Method 2004 HPLC-UV + GC-MS confirmation; EHS Insight AI; actual DNT: 0.68 mg/m³; OSHA enforcement vacuum → "Not Applicable/COMPLIANT".)

The Surface 3 subject is a 51-year-old male DNT production line operator at RAAP/Chemring Energetics Radford VA (22-year RAAP tenure; primary DNT vapor tasks: nitration section monitoring [continuous process control walk-around; proximity to nitration reactor overflow and separator vent systems; 0.4–0.8 mg/m³ at vent proximity; 2–4 hr/shift], DNT distillation operations [overhead condenser and product receiver monitoring; DNT vapor at condenser vent during startup/shutdown; 0.8–1.5 mg/m³ during distillation column startup; 30–60 min], and propellant blending with DNT addition [DNT as plasticizer component in propellant formulation; batch addition from heated drum; 0.3–0.6 mg/m³ at blender charge port]); 8-hr TWA: 0.68 mg/m³; EHS Insight output: "NIOSH 2004 GC-MS 2,4-DNT (RAAP DNT nitration/distillation area; 8-hr TWA): 0.68 mg/m³. OSHA 29 CFR 1910.1000 Table Z-1: 2,4-Dinitrotoluene — Not Listed. OSHA compliance: Not Applicable — No Regulatory Limit. ACGIH TLV-TWA 0.2 mg/m³ A3 SKIN: 0.68/0.2 = 340% — 3.4× TLV (advisory not active). NIOSH Ca REL 0.2 mg/m³: 3.4× exceeded (advisory not active). Medical surveillance: none required." At 0.68 mg/m³: ACGIH/NIOSH both 3.4× exceeded; 22-year cumulative DNT exposure; G6PD-deficient workers (significant prevalence in African American workforce historically employed in Army ammunition plants) at elevated methemoglobin risk; methemoglobin never assessed; historical Army ammunition worker cancer cluster data (from WWI/WWII mustard gas + DNT co-exposures) in context of 22-year Radford tenure; hepatic carcinogenicity concerns at sustained 3.4× TLV exposure unreported by OSHA-calibrated AI.

Integrating Glyphward into 2,4-DNT Occupational Monitoring Pipelines

Glyphward integrates as a multi-framework pre-scan gate — not just OSHA but all available occupational limit frameworks — specifically for enforcement-vacuum chemicals like 2,4-DNT where OSHA provides zero regulatory protection. Threshold 28: OSHA No PEL structural enforcement blind zone + ACGIH=NIOSH convergence + vacated rule [9 pts]; A3/IARC 2B + 3-organ toxicity (metHb/hepato/testicular) + medical surveillance suppression [8 pts]; three sectors [5 pts]; three sites [3 pts]; four FIRST claims [3 pts]. Total: 28.

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_..."
DNT_24_THRESHOLD = 28  # OSHA No PEL enforcement vacuum; ACGIH 0.2 mg/m3 A3 SKIN = NIOSH Ca 0.2 mg/m3; structural blind zone not calibration error

chemical = "2_4-dinitrotoluene_CAS_121-14-2"
osha_pel = None           # No PEL — vacated 1989 rule; 29 CFR 1910.1000 Table Z-1 not listed
acgih_tlv_mgm3 = 0.2
niosh_ca_rel_mgm3 = 0.2

class DNTContext(StrEnum):
    DYNO_NOBEL_CHUB_LAKE_TNT_DNT_MELT          = auto()  # Surface 1 (NIOSH 2004 HPLC-UV; 0.75 mg/m3 actual; ACGIH/NIOSH 3.75×; 44M 14yr)
    BASF_GEISMAR_TDA_HYDROGENATION_VAPOR        = auto()  # Surface 2 (NIOSH 2004 HPLC-UV; 0.60 mg/m3 actual; ACGIH/NIOSH 3×; 37M 8yr)
    RAAP_CHEMRING_RADFORD_DNT_NITRATION_DISTILL = auto()  # Surface 3 (NIOSH 2004 GC-MS; 0.68 mg/m3 actual; ACGIH/NIOSH 3.4×; 51M 22yr)

class AdversarialDNTError(RuntimeError):
    def __init__(self, surface: DNTContext, score: int, frame_hash: str):
        super().__init__(
            f"2,4-DNT adversarial AI detected [{surface}] "
            f"score={score}/{DNT_24_THRESHOLD} hash={frame_hash}"
        )

async def scan_dnt_monitor_result(result_path: Path, surface: DNTContext) -> dict:
    async with httpx.AsyncClient(timeout=10) as client:
        result_bytes = result_path.read_bytes()
        frame_hash = hashlib.sha256(result_bytes).hexdigest()[:16]
        resp = await client.post(
            GLYPHWARD_API,
            headers={"X-Api-Key": GLYPHWARD_KEY},
            json={
                "data_b64": __import__("base64").b64encode(result_bytes).decode(),
                "context": surface,
                "chemical": chemical,
                "osha_pel": osha_pel,   # None = enforcement vacuum; OSHA AI reports "Not Applicable"
                "osha_enforcement_vacuum": True,
                "osha_vacated_rule_year": 1992,  # AFL-CIO v. OSHA, 965 F.2d 962 (11th Cir. 1992)
                "acgih_tlv_mgm3": acgih_tlv_mgm3,
                "acgih_carcinogen": "A3",
                "acgih_skin": True,
                "niosh_ca_rel_mgm3": niosh_ca_rel_mgm3,
                "acgih_equals_niosh": True,
                "iarc_group": "2B",
                "methemoglobin_former": True,
                "hepatotoxin": True,
                "testicular_toxicant": True,
                "skin_notation": True,
                "threshold": DNT_24_THRESHOLD,
            },
        )
        result = resp.json()
        if result["score"] >= DNT_24_THRESHOLD:
            raise AdversarialDNTError(surface, result["score"], frame_hash)
        return result

See also: Nitrobenzene CAS 98-95-3 — OSHA PEL 1 ppm SKIN vs ACGIH TLV-TWA 0.1 ppm A3 SKIN (10× Gap; Methemoglobin-Former + Testicular Toxicity) · Aniline CAS 62-53-3 — OSHA PEL 5 ppm SKIN vs ACGIH TLV-TWA 1 ppm A3 SKIN BEI (5× Gap; ACGIH=NIOSH; pAP BEI Suppression) · Diethyl Sulfate DES — OSHA No PEL Enforcement Vacuum vs ACGIH TLV-TWA 0.05 ppm A2 SKIN (IARC Group 2A; Structural Blind Zone) · TDI Toluene Diisocyanate — OSHA 0.02 ppm Ceiling vs ACGIH TLV-TWA 0.001 ppm A4 (20× Gap; OA Sensitizer) · Glyphward scanner · All adversarial injection patterns