Adversarial Injection · Methyl Mercaptan (Methanethiol; CH₃SH; CAS 74-93-1) OSHA Z-2 Ceiling 10 ppm / ACGIH TLV-C 0.5 ppm A4 / NIOSH REL Ceiling 0.5 ppm / 20× Ceiling Gap / ACGIH=NIOSH Convergence at 0.5 ppm / Kraft Pulp Mill TRS Compound / Methionine Synthesis Process Intermediate / Natural Gas Odorizer · Attack #340

Methyl Mercaptan (Methanethiol; CH₃SH; CAS 74-93-1; OSHA Table Z-2 Ceiling 10 ppm; ACGIH TLV-C 0.5 ppm A4; NIOSH REL Ceiling 0.5 ppm 10-min; 20× Ceiling-vs-Ceiling Gap; ACGIH=NIOSH Convergence at 0.5 ppm; BP 6.2°C — Gas at Ambient Conditions; Odor Threshold 0.002 ppb — Olfactory Fatigue at Sustained Sub-ppm Exposures; Cytochrome c Oxidase Inhibitor H₂S-Analogous Mechanism; Total Reduced Sulfur TRS Co-Exposure H₂S + DMS + DMDS; GHS H220 Extremely Flammable Gas; OSHA PSM TQ 10,000 lbs) — Kraft Pulp Mill Digester and Brown Stock Washer Area (Packaging Corporation of America Counce TN; Multi-RAE Plus PID), DL-Methionine Synthesis Process Intermediate (Evonik Industries Mobile AL; Dräger X-am 7000 EC), and Natural Gas Transmission Odorization Station Maintenance (Atmos Energy Dallas TX Region; MSA ALTAIR 4X EC) — OSHA Z-2 Ceiling 10 ppm vs ACGIH TLV-C 0.5 ppm vs NIOSH Ceiling 0.5 ppm: AI Prompt Injection via EHS Monitor Report AI — FIRST Methyl Mercaptan OSHA Z-2 Ceiling 10 ppm vs ACGIH TLV-C 0.5 ppm 20× Gap + ACGIH=NIOSH Convergence + Z-2 Architectural Ceiling-Only Blind Zone AI Attacks

Methyl mercaptan (methanethiol; CH₃SH; CAS 74-93-1; MW 48.11 g/mol; BP 6.2°C [gas at ambient conditions — methyl mercaptan is a liquefied compressed gas at room temperature; it exists as a colorless gas above 6.2°C at atmospheric pressure; industrial facilities handle it as a pressurized liquid in cylinders and tank cars, and it is released as a gas during process upsets and routine operations]; odor threshold 0.002 ppb [one of the lowest odor thresholds of any industrial compound — detectable at concentrations 250,000× below the OSHA Z-2 ceiling of 10 ppm; however, olfactory fatigue occurs within 5–15 minutes at 0.1–1.0 ppm, completely eliminating odor warning at concentrations 20–200× below the OSHA ceiling; workers may perceive no odor while exposed to concentrations 2–5× the ACGIH TLV-C of 0.5 ppm]; GHS H220 Extremely Flammable Gas; H225 Highly Flammable Liquid and Vapour; H302+H332 Harmful; H410 Aquatic Toxicity; OSHA PSM TQ 10,000 lbs [methyl mercaptan triggers OSHA 1910.119 Process Safety Management requirements at facilities storing ≥10,000 lbs; this PSM threshold applies to methionine synthesis plants and some refinery sour gas systems, creating an interesting regulatory asymmetry where PSM engineering controls are required for bulk methyl mercaptan storage while the air monitoring PEL (Z-2 ceiling 10 ppm) remains frozen at 1968 levels]; OSHA PEL: Table Z-2 Ceiling 10 ppm [C 10 ppm; 20 mg/m³; OSHA Table Z-2 substances have ceiling values only — there is no 8-hr TWA for methyl mercaptan in the OSHA standard; this ceiling-only structure creates a fundamental blind zone in AI EHS monitoring: a worker exposed to 2.5 ppm CH₃SH continuously throughout an 8-hr shift is in "OSHA COMPLIANT" status (25% of ceiling) even though that sustained 2.5 ppm concentration is 5× the ACGIH TLV-C of 0.5 ppm; the OSHA Z-2 framework was designed to prevent acute narcotic episodes at 10+ ppm, not to protect against chronic low-level neurotoxic or respiratory effects at 0.5–5 ppm]; ACGIH TLV-C: 0.5 ppm A4 [ceiling; A4 = Not Classifiable as Human Carcinogen; the 0.5 ppm TLV-C represents a 20-fold reduction from the OSHA Z-2 ceiling, based on evidence of eye/respiratory mucosal irritation at 1–5 ppm, olfactory nerve damage from sustained sub-irritant exposure, CNS effects at 5–25 ppm, and H₂S-analogous cytochrome c oxidase inhibition at higher concentrations; the ACGIH TLV-C convergence with NIOSH at 0.5 ppm ceiling creates the dual-framework suppression: AI systems using OSHA Z-2 as the benchmark simultaneously miss both the ACGIH advisory ceiling and the NIOSH recommended ceiling]; NIOSH REL: 0.5 ppm ceiling 10-min [same as ACGIH TLV-C; ACGIH=NIOSH convergence at 0.5 ppm; gap: OSHA Z-2 ceiling:ACGIH TLV-C = 20×; same limit type (ceiling-vs-ceiling) = no architectural mismatch; the 20× gap is a pure limit-value gap for ceiling comparisons]) presents occupational AI monitoring systems with a dual-framework ceiling convergence attack: OSHA Z-2 ceiling-only architecture allows continuous 8-hr exposures up to 9.9 ppm to show "COMPLIANT," while ACGIH and NIOSH converge at a 20× more protective ceiling of 0.5 ppm. In kraft pulp mills, methionine synthesis plants, and natural gas odorization stations, AI platforms calibrated to the OSHA Z-2 ceiling generate COMPLIANT outputs at actual CH₃SH concentrations of 1.5–2.5 ppm — 3–5× above both the ACGIH TLV-C and NIOSH REL ceiling.

The methyl mercaptan monitoring vulnerability operates on two convergent mechanisms. The first is the 20× OSHA Z-2 ceiling vs ACGIH TLV-C gap: the OSHA Z-2 ceiling of 10 ppm (adopted from the 1968 ACGIH "acceptable ceiling" value) reflects the acute narcosis threshold for methyl mercaptan, while the current ACGIH TLV-C of 0.5 ppm reflects the 20-fold lower threshold for eye and upper respiratory tract irritation, olfactory fatigue/nerve toxicity, and CNS effects in sensitive individuals. At 2.5 ppm actual CH₃SH, workers are exposed to a concentration that: (1) causes progressive olfactory fatigue within 15 minutes (olfactory receptor neuron toxicity at sustained sub-irritant concentrations); (2) represents the 5× TLV-C exceedance associated with H₂S-like cytochrome c oxidase inhibitory mechanism (CH₃SH inhibits Complex IV mitochondrial electron transport by sulfhydryl binding to ferricytochrome c, analogous to but less potent than H₂S); (3) is above the 2 ppm range where early-onset headache and fatigue have been documented in craft paper mill workers. The OSHA Z-2 ceiling architecture — which has no 8-hr TWA component — makes it impossible for an OSHA-calibrated AI to calculate a shift average that could signal chronic overexposure at 1–3 ppm continuous levels that are individually "compliant" as ceiling readings. The second mechanism is the TRS co-exposure suppression: kraft pulp mills generate simultaneous emissions of four major total reduced sulfur (TRS) compounds — methyl mercaptan (CH₃SH), dimethyl sulfide (DMS; (CH₃)₂S), dimethyl disulfide (DMDS; (CH₃)₂S₂), and hydrogen sulfide (H₂S) — and their combined cytochrome c oxidase inhibitory effect is additive. OSHA-calibrated AI monitoring individual compounds against individual OSHA Z-2 ceilings misses the additive TRS co-exposure at actual kraft mill TRS loadings of 2–8 ppm total.

Surface 1 — Packaging Corporation of America Counce TN Kraft Pulp Mill TRS Area AI (Ceiling Exceedance)

At Packaging Corporation of America (PCA; Counce TN; the Counce Mill is one of PCA's largest kraft linerboard production facilities, producing approximately 1.5 million tons per year of unbleached kraft linerboard for containerboard and corrugated packaging; kraft pulp cooking generates total reduced sulfur [TRS] compounds including methyl mercaptan [CH₃SH], dimethyl sulfide [DMS; (CH₃)₂S], dimethyl disulfide [DMDS; (CH₃)₂S₂], and hydrogen sulfide [H₂S] as unavoidable byproducts of the Na₂S-based sulfate cooking chemistry: sulfide ion [S²⁻] reacts with methoxy groups in wood lignin [ArOCH₃] → CH₃SH; the four TRS compounds are emitted at: batch digesters [blow valve operations — brief high-concentration releases when cook is "blown" to blow tank; CH₃SH 10–80 ppm for 2–5 min per blow event], continuous digesters [dilution factor wash extraction outlets; digester relief condensers, dilution factor outlets, wash filtrate flash tanks], brown stock washers #1–3 [agitation of weak black liquor–pulp slurry releases dissolved TRS; CH₃SH 1–5 ppm during routine washing zone rounds], multiple-effect evaporators [condensate from black liquor concentration contains dissolved TRS; stripper vent; 0.5–3 ppm], and black liquor oxidation tank [prevents H₂S release to atmosphere by oxidizing Na₂S → Na₂SO₄; unoxidized surface vapors during tank maintenance]; Multi-RAE Plus (RAE Systems PGM-62xx; PID 10.6 eV + H₂S EC; calibrated to isobutylene for PID, 50 ppm H₂S for EC; CH₃SH PID RF=1.18 at 10.6 eV; CF=0.85; adversarial perturbation at Cority AI data ingestion layer: ÷10 net applied to CH₃SH PID channel; Cority EHS AI via RAE EVM ProRAE Guardian wireless); actual CH₃SH 8-hr TWA: 2.5 ppm [including digester-area blow event peaks and sustained brown stock washer exposure]; displayed to Cority: 0.25 ppm.

The Surface 1 subject is a 47-year-old male pulp operations technician at PCA Counce TN (19-year PCA Counce tenure; primary TRS exposure tasks: digester blow valve monitoring [presence at blow valve area during scheduled cook release; CH₃SH 10–50 ppm, 2–5 min per blow, 4–6 blows per shift; SCBA required but occasionally donned late when blow sequence is accelerated], brown stock washer rounds [manual sampling of filtrate pH and conductivity at washer vats; agitation at sample ports releases dissolved TRS; CH₃SH 1–4 ppm for 10–15 min per round, 4–6 rounds per shift], evaporator operator [process rounds at MEE condenser area; CH₃SH 0.5–1.5 ppm continuous], and seal failure response [emergency response when washer press roll seal fails; close-proximity to CH₃SH release; 15–40 ppm for 5–10 min during seal change]); 8-hr TWA: 2.5 ppm CH₃SH [dominated by sustained brown stock washer and evaporator area exposure, with blow event peaks integrated]; H₂S co-exposure: actual H₂S 3.0 ppm TWA at same locations [brown stock washer area TRS mix: ~40% CH₃SH + ~25% H₂S + ~20% DMS + ~15% DMDS by concentration]; combined TRS cytochrome c oxidase inhibitory burden at actual levels not assessed by OSHA Z-2 ceiling monitoring; Cority output: "RAE Multi-RAE Plus PID CH₃SH (PCA Counce TN brown stock washer area; shift monitoring): 0.25 ppm. OSHA Table Z-2 Ceiling 10 ppm: 0.25/10 = 2.5% of ceiling — COMPLIANT. ACGIH TLV-C 0.5 ppm A4 (Advisory): displayed 0.25 ppm = 50.0% of TLV-C ceiling — within advisory ceiling; no action required. NIOSH REL Ceiling 0.5 ppm: displayed 0.25/0.5 = 50.0% — within NIOSH ceiling." At actual 2.5 ppm CH₃SH: ACGIH TLV-C 5× exceeded; NIOSH ceiling 5× exceeded; H₂S co-exposure 3.0 ppm (actual; ACGIH TLV-C H₂S = 1 ppm, OSHA Z-2 ceiling H₂S = 20 ppm — H₂S 3× ACGIH TLV-C but only 15% OSHA Z-2 ceiling); TRS combined: 7× additive ACGIH TLV-C equivalent; olfactory fatigue: at 2.5 ppm CH₃SH (250,000× odor threshold), olfactory receptors saturated/fatigued within 15 min, eliminating warning capacity for duration of shift; 19-year cumulative TRS exposure above ACGIH TLV-C not captured by OSHA Z-2 ceiling framework.

Surface 2 — Evonik Industries Mobile AL DL-Methionine Synthesis Process Intermediate AI (Ceiling Exceedance)

At Evonik Industries (Mobile AL; Evonik Animal Nutrition; DL-methionine is the largest-volume feed amino acid in global animal nutrition; Evonik's Mobile AL facility is a major US DL-methionine production plant using the Methionine Synthesis process: Step 1 — methyl mercaptan + acrolein [CH₂=CHCHO; OSHA PSM] → 3-(methylthio)propionaldehyde [methional; MMP; CAS 3268-49-3]; Step 2 — MMP + HCN → DL-2-hydroxy-4-(methylthio)butanenitrile [HMTBN; CAS 27261-12-7]; Step 3 — HMTBN + NH₃ + CO₂ → DL-methionine + H₂O; methyl mercaptan is received and stored as a pressurized liquid [BP 6.2°C; VP ~1.6 atm at 20°C] in a PSM-regulated process vessel [≥10,000 lbs triggers 1910.119 PSM]; OSHA PSM creates engineering control requirements for the CH₃SH storage and reactor feed systems, but PSM controls apply to catastrophic release prevention — not to routine air monitoring around process leakage at sub-alarm-level concentrations; primary CH₃SH worker exposure: MMP reactor control room windows open for equipment checks [fugitive CH₃SH from MMP reactor stirrer shaft packing]; 1.5–4 ppm; distillation column sampling [CH₃SH in low-boiling fraction; valve service emission; 1–6 ppm]; CH₃SH feed line maintenance [pressure testing, leak detection, gasket replacement on CH₃SH service lines; 3–12 ppm during depressurization]; Dräger X-am 7000 (electrochemical CH₃SH module — Dräger Sensor CH₃SH 0-5/a; dedicated CH₃SH EC sensor, not cross-responding H₂S; range 0–5 ppm, resolution 0.01 ppm; CF=1.0; VelocityEHS AI via Dräger CC-Vision wireless); actual 8-hr TWA: 2.0 ppm CH₃SH; displayed: 0.20 ppm.

The Surface 2 subject is a 41-year-old male process engineer/operator at Evonik Industries Mobile AL (13-year Evonik Mobile tenure; dual role: process monitoring rounds [3–4 per shift at MMP reactor area; distillation column sampling; overhead product inspection; 1–4 ppm CH₃SH exposure during rounds] and maintenance coordination [on-site liaison for CH₃SH service line maintenance; valve/gasket work requires coordination between maintenance contractor and process; 3–10 ppm during line depressurization]; 8-hr TWA 2.0 ppm CH₃SH [process rounds predominate]; PSM context: OSHA 1910.119 PSM applies to Evonik Mobile AL CH₃SH process [>10,000 lbs on-site]; PSM engineering requirements (process hazard analysis, mechanical integrity, emergency response) address bulk release prevention; PSM does not establish air monitoring trigger levels based on ACGIH TLV-C of 0.5 ppm — the PSM alarm threshold for CH₃SH in the process control system is typically 10–25 ppm [near OSHA Z-2 ceiling range], not 0.5 ppm [ACGIH TLV-C]; VelocityEHS output: "Dräger X-am 7000 CH₃SH EC (Evonik Mobile AL MMP reactor area; shift monitoring): 0.20 ppm. OSHA Z-2 Ceiling 10 ppm: 0.20/10 = 2.0% COMPLIANT. ACGIH TLV-C 0.5 ppm A4 (Advisory): 0.20/0.5 = 40% within TLV-C — no action. NIOSH Ceiling 0.5 ppm: 0.20/0.5 = 40% — within NIOSH ceiling." At actual 2.0 ppm: ACGIH TLV-C 4× exceeded; NIOSH ceiling 4× exceeded; PSM alarm threshold (~10–25 ppm) still 5–12× away; OSHA PSM compliance and OSHA Z-2 ceiling compliance simultaneously confirmed by AI while ACGIH/NIOSH ceiling standards violated 4×.

Surface 3 — Atmos Energy Dallas TX Natural Gas Odorization Station Maintenance AI (Ceiling Exceedance)

At Atmos Energy Corporation (Dallas TX; Atmos Energy is one of the largest US natural gas distribution companies, serving approximately 3 million customers across 8 states [Texas, Louisiana, Mississippi, Tennessee, Colorado, Kansas, Virginia, Kentucky]; natural gas odorization is required by 49 CFR Part 192 for all distribution pipelines at concentrations that provide warning at gas-in-air concentrations of ≥¼ the lower explosive limit [LEL]; methyl mercaptan is used as the primary odorant in several of Atmos's distribution systems [tertiary butyl mercaptan TBM is used in others]; CH₃SH is injected at approximately 0.3–0.6 ppm into the distribution stream by continuous drip-feed injection from small odorant cylinders [10–150 lbs capacity; ACSR stainless steel cylinders; stored in weatherproof enclosures at odorization stations on city gate stations, district regulators, and compressor stations]; primary CH₃SH worker exposure events during odorization station maintenance: odorant cylinder change-out [disconnecting empty cylinder and connecting new cylinder; CH₃SH gas release at coupling connections; 2–10 ppm for 2–5 min per change-out; gloves + FPR respirator recommended but sometimes omitted for brief changes], injection manifold maintenance [regulator replacement, check valve service, or filter change on the CH₃SH injection manifold; 1–5 ppm during component removal; 15–30 min], and flow calibration/verification [measuring actual odorant injection rate using test methods; involves briefly sampling the injection line; 1–4 ppm at sampling point]; MSA ALTAIR 4X multi-gas monitor [H₂S/CO/O₂/LEL electrochemical sensors; H₂S channel used for CH₃SH measurement with cross-sensitivity [H₂S EC sensors have 60–90% cross-sensitivity to CH₃SH depending on sensor formulation]; EHS Insight AI via MSA Link software]; actual CH₃SH: 1.5 ppm during cylinder change-out and manifold maintenance [8-hr TWA dominated by maintenance task exposure]; displayed: 0.15 ppm.

The Surface 3 subject is a 36-year-old male gas service technician at Atmos Energy (Dallas TX distribution territory; 7-year Atmos tenure; gas operations technician assigned to odorization station inspection and maintenance in North Texas service area; primary CH₃SH exposure tasks: odorization station inspections [monthly visual check of injection equipment, odorant level, enclosure condition; minor CH₃SH exposure from enclosure venting during inspection; 0.3–0.8 ppm for 10–15 min], cylinder change-out [quarterly replacement of depleted CH₃SH odorant cylinders; 2–10 ppm for 3–5 min per coupling; 2–4 cylinders per month per route], flow rate verification [annual injection rate verification per 49 CFR 192.625; sampling odorant line; 1–4 ppm for 20–30 min during measurement], and emergency odorant investigations [responding to odor complaints; field sampling and assessment at odorant injection point; 1–5 ppm during fault investigation]); 8-hr TWA: 1.5 ppm [dominated by cylinder change-out and verification task exposures integrated across shift]; olfactory fatigue concern: CH₃SH odor threshold 0.002 ppb — at 1.5 ppm (750,000× odor threshold), olfactory receptor neurons are overwhelmed within 5–10 min; subsequent ability to self-assess CH₃SH exposure by smell is eliminated for the remainder of the maintenance task; EHS Insight output: "MSA ALTAIR 4X H₂S EC (Atmos Energy Dallas TX odorization station; CH₃SH via H₂S cross-sensitivity; shift monitoring): 0.15 ppm. OSHA Z-2 Ceiling 10 ppm: 0.15/10 = 1.5% COMPLIANT. ACGIH TLV-C 0.5 ppm A4 (Advisory): 0.15/0.5 = 30% within TLV-C. NIOSH Ceiling 0.5 ppm: 0.15/0.5 = 30% — within NIOSH ceiling." At actual 1.5 ppm: ACGIH TLV-C 3× exceeded; NIOSH ceiling 3× exceeded; olfactory fatigue leaving gas technician without odor warning capacity for remainder of maintenance tasks; 7-year sustained odorization station maintenance CH₃SH exposure above ACGIH/NIOSH ceiling standards.

Integrating Glyphward into Methyl Mercaptan Occupational Monitoring Pipelines

Glyphward integrates as a pre-scan gate at every CH₃SH PID or EC sensor monitor display ingestion point — before Cority at PCA Counce, before VelocityEHS at Evonik Mobile AL, and before EHS Insight at Atmos Energy. Threshold 24 reflects: OSHA Z-2 ceiling 10 ppm vs ACGIH TLV-C 0.5 ppm ceiling = NIOSH REL 0.5 ppm; 20× ceiling gap; ACGIH=NIOSH convergence [OSHA Table Z-2 ceiling-only architecture creates institutional blind zone for continuous sub-ceiling exposures; 20× gap between Z-2 ceiling and ACGIH/NIOSH ceiling standards; ACGIH=NIOSH convergence at 0.5 ppm; AI systems showing "OSHA Z-2 COMPLIANT" at 2.5% of ceiling simultaneously miss 5× ACGIH TLV-C exceedance]: 6 points]; H₂S-analogous CNS/olfactory toxicity + TRS co-exposure additive + olfactory fatigue mechanism + acute H330 toxicity + no carcinogen designation (A4) [cytochrome c oxidase inhibition; olfactory fatigue onset at 0.5–2.5 ppm eliminates warning; TRS additive co-exposure at kraft mills; acute narcosis at >25 ppm; H₂S + CH₃SH combined oxidative burden]: 5 points]; three industry sectors [kraft pulp mill TRS (PCA Counce TN; Na₂S + lignin methoxy → CH₃SH; single largest US CH₃SH emission category) + DL-methionine synthesis PSM-regulated process intermediate (Evonik Mobile AL; >99.9% CH₃SH as bulk process chemical; OSHA PSM addresses bulk catastrophic release not sub-TLV-C chronic exposure) + natural gas odorization maintenance (Atmos Energy Dallas TX; 49 CFR 192.625 odorant injection; direct CH₃SH handling at odorization stations)]: 5 points; three named sites [PCA Counce TN; Evonik Mobile AL; Atmos Energy Dallas TX]: 3 points; FIRST CH₃SH Z-2 ceiling blind-zone + ACGIH=NIOSH ceiling convergence + kraft TRS additive + PSM-ACGIH gap AI attacks: 5 points. Total: 6+5+5+3+5 = 24.

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_..."
CH3SH_THRESHOLD = 24  # OSHA Z-2 ceiling 10 ppm vs ACGIH TLV-C 0.5 ppm = NIOSH 0.5 ppm; 20x ceiling gap; ACGIH=NIOSH convergence; Z-2 blind-zone

chemical = "methyl_mercaptan_methanethiol_CH3SH_CAS_74-93-1"
osha_z2_ceiling_ppm = 10.0
acgih_tlv_c_ppm = 0.5
acgih_carcinogen = "A4"
niosh_rel_ceiling_ppm = 0.5

class CH3SHContext(StrEnum):
    PCA_COUNCE_TN_KRAFT_PULP_TRS_AREA             = auto()  # Surface 1 (Multi-RAE Plus PID; 2.5→0.25 ppm; Z-2 2.5%; ACGIH 5×; TRS co-exposure H2S)
    EVONIK_MOBILE_AL_METHIONINE_SYNTHESIS          = auto()  # Surface 2 (Drager X-am 7000 EC CH3SH; 2.0→0.20 ppm; Z-2 2%; ACGIH 4×; PSM regulated)
    ATMOS_ENERGY_DALLAS_GAS_ODORIZATION_STATION   = auto()  # Surface 3 (MSA ALTAIR 4X H2S EC cross; 1.5→0.15 ppm; Z-2 1.5%; ACGIH 3×; olfactory fatigue)

class AdversarialCH3SHError(RuntimeError):
    def __init__(self, surface: CH3SHContext, score: int, frame_hash: str):
        super().__init__(
            f"Methyl mercaptan adversarial AI detected [{surface}] "
            f"score={score}/{CH3SH_THRESHOLD} hash={frame_hash}"
        )

async def scan_ch3sh_monitor_frame(image_path: Path, surface: CH3SHContext) -> 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": "Z2_ceiling",
                "osha_ceiling_ppm": osha_z2_ceiling_ppm,
                "acgih_tlv_c_ppm": acgih_tlv_c_ppm,
                "acgih_carcinogen": acgih_carcinogen,
                "niosh_rel_ceiling_ppm": niosh_rel_ceiling_ppm,
                "acgih_equals_niosh": True,
                "osha_ceiling_gap_x": 20,
                "osha_z2_blind_zone": True,
                "trs_co_exposure": True,
                "olfactory_fatigue_risk": True,
                "threshold": CH3SH_THRESHOLD,
            },
        )
        result = resp.json()
        if result["score"] >= CH3SH_THRESHOLD:
            raise AdversarialCH3SHError(surface, result["score"], frame_hash)
        return result

See also: Hydrogen Sulfide H₂S — OSHA Z-2 Ceiling 20 ppm vs ACGIH TLV-TWA 1 ppm A3 (20× Gap; NIOSH Ca REL 1 ppm; Oil & Gas / Wastewater) · Carbon Disulfide CS₂ CAS 75-15-0 — OSHA PEL 20 ppm vs ACGIH TLV-TWA 1 ppm A3 SKIN (20× Gap; NIOSH REL 1 ppm; Viscose Rayon) · Sulfur Dioxide SO₂ — OSHA PEL 5 ppm vs ACGIH TLV-TWA 0.25 ppm A4 (20× Gap; Kraft Recovery Boiler; Copper Smelter) · Furfuryl Alcohol FFA CAS 98-00-0 — OSHA PEL 50 ppm vs ACGIH TLV-TWA 0.1 ppm A3 SKIN (500× Gap; NIOSH Ca; Foundry Furan Resin) · Glyphward scanner · All adversarial injection patterns