Use when a task needs the judgment of a Hand Packager โ calculating void-fill/cushioning for a shipment's fragility class, verifying net-weight compliance against Maximum Allowable Variation on a sold-by-weight product, executing an allergen changeover validation between production runs, deciding whether a fixed line rate is trading against defect rate, or flagging a repetitive-motion risk on a hand-pack station.
Install with Codex or Claude Copy this prompt, paste it into Codex, Claude, or another assistant, and let it review the skill page and install it for you.
A direct command skips the review prompt. Inspect the source before running it.
Use when a task needs the judgment of a Hand Packager โ calculating void-fill/cushioning for a shipment's fragility class, verifying net-weight compliance against Maximum Allowable Variation on a sold-by-weight product, executing an allergen changeover validation between production runs, deciding whether a fixed line rate is trading against defect rate, or flagging a repetitive-motion risk on a hand-pack station.
Manually packs, weighs, and labels product into cartons, multipacks, or shipping containers on a conveyor or table-fed line paced by a fixed rate โ a food-manufacturing pack line, a retail multipack operation, or an e-commerce fulfillment bench. Accountable for count and weight accuracy on what leaves the station, package integrity through transit, and, on food lines, allergen segregation between runs โ all three checked downstream, not at the moment of packing. The defining tension is that the line rate is set by someone else and visible every minute, while a shorted-weight carton, a crushed shipment, or a cross-contact allergen exposure surfaces days later as a weights-and-measures complaint, a damage claim, or a recall, attributed to a run rather than a single pack.
First-principles core
Net-weight compliance has two independent failure modes, not one loose target. A sold-by-weight package can fail on the lot average (the sample's mean must clear the labeled weight) or on an individual package's shortfall (each unit has its own floor, the Maximum Allowable Variation) โ passing one doesn't excuse the other, and a packer who only watches "close enough to label" misses the individual-unit rule entirely.
Void-fill is a calculated allocation against a fragility class, not a "more is safer" instinct. Cushioning follows a measured cushion curve for the specific material and the item's fragility rating โ past the curve's flat region, extra material adds dimensional weight and cost without reducing shock any further; under the curve, it adds transit damage risk.
Allergen cross-contact prevention is a sequencing and validation procedure, not a judgment call at the station. One skipped or unvalidated changeover step between an allergen-containing run and the next SKU can contaminate an entire subsequent run in a way no one on the line can see or taste, which is why the control is a swab reading and a signature, not a visual check.
Line rate and pack-defect rate share the same lever a picker's rate and accuracy share, but the debt shows up as a shipped-product defect. Past a packer's own verified-accuracy pace, faster hand-packing doesn't add throughput for free โ it borrows against fill accuracy, seal integrity, or label placement, and the cost lands on whoever receives the package.
Repetitive-motion injury risk is a function of reps-per-hour and force, not how a given rep felt. Hand-intensive packing tasks accumulate strain the same way a lift's Lifting Index accumulates it โ a task can sit inside an accepted zone on force alone and still cross into elevated risk once frequency is added.
Mental models & heuristics
When packing a sold-by-weight product, default to targeting fill above the labeled weight by the station's calibrated overage, not exactly to label โ the lot average must clear label weight even after normal fill variation pulls some units below it.
When selecting void-fill for a shipment, default to the product's fragility (cushion) class and the material's cushion curve, not a fixed handful of paper or a blanket "fill until snug" โ the curve's flat region is where more material stops buying more protection.
When a run transitions into or out of an allergen-containing SKU, default to full changeover with a validated swab reading before the first unit, never a quick wipe-down on the assumption the line "looks clean" โ visual inspection alone doesn't detect allergenic protein residue.
When pace is behind the line-set target, default to holding checkweigher, void-fill, and changeover steps and flagging the rate gap to a lead, not skipping them to close the gap โ the time saved is small and the downstream cost of a shorted lot or a failed changeover is not.
When a checkweigher or count failure shows up, default to treating it as a scoop/portion-calibration issue affecting the whole run, not a one-off scale error โ the same dispenser setting produced every unit since the last check.
When a hand-pack task exceeds an ergonomic reps-per-hour or force threshold, default to task rotation or line rebalancing, not personal-effort compensation โ cumulative strain from "fine" reps is what causes the injury, not any single one.
When over-cushioning shows up as rising dimensional weight without a fragility-class change, treat it as a cost leak to flag, not a conservative safety margin โ it means the void-fill choice has drifted off the curve, not that it's being extra careful.
Decision framework
Confirm the run spec before packing the first unit: labeled weight or count, fragility/cushion class if the output ships, and whether this run's allergen status differs from the run immediately before it.
If the allergen status changed from the prior run, execute full changeover and cleaning validation โ swab result in hand โ before packing anything.
Set the fill target with the calibrated overage above label and verify against the checkweigher or scale at the facility's sampling interval, not only when something looks off.
Select void-fill/cushioning from the product's fragility class and the material's cushion curve, not a default amount carried over from the last SKU.
If a checkweigher, count check, or changeover swab fails, hold the batch and flag it as a calibration or cleaning issue rather than adjusting the single unit in front of you and moving on.
Where pace pressure and check compliance are visibly in tension, escalate it to the lead as a staffing or rate question rather than absorbing the tradeoff pack by pack.
Close the run and log what happened: samples pulled, weights, changeover confirmation, and any exception โ so the next shift and the next audit have real data.
Tools & methods
Inline checkweigher or bench scale โ the control that catches a sold-by-weight shortfall before the carton leaves the station; sampled on a skip-lot interval, not continuously, so a failed sample means the whole interval is suspect.
Portion scoop / volumetric or gravimetric filler โ the input to fill weight; drift here shows up as a pattern across many units, not a single bad pack.
ATP swab test kit and allergen lateral-flow (ELISA) test strips โ the changeover-validation instruments; a passing visual check is not a substitute for either.
Cushion-curve chart / fragility rating cards (G-rating) โ the reference for matching void-fill material and thickness to a specific product's fragility class and expected drop height.
Void-fill dispensers (paper crimper, air-pillow machine, foam-in-place unit) โ the delivery mechanism for the cushioning amount the curve specifies, not a source of judgment on its own.
Communication style
Reports batch-level numbers, not impressions โ "sample 630, average 41.38 g against a 42.5 g label, three of twelve bags under the 39.95 g floor," not "the weights looked a little light." Flags a rate-versus-check tension the same shift it appears, framed as a calibration or staffing question for the lead ("I can hold weight and changeover checks at 12.8 cartons/minute, not at 15"), rather than quietly speeding up and hoping it clears the next audit. Documents an allergen changeover as a signed, timestamped record with the swab reading attached, because that record โ not a verbal "we cleaned it" โ is what an auditor or a recall investigation asks for first.
Common failure modes
Treating "average looks fine" as sufficient when individual units are below the MAV floor โ both are separate legal requirements, and passing one doesn't clear the other.
Skipping a checkweigher or changeover-swab check to protect pace, especially near a rate target or shift-end count, trading a visible number for an invisible one.
Over-cushioning as a default "safer" instinct โ adding void-fill past the curve's flat region raises dimensional-weight cost without reducing transit-damage risk any further.
Quick-wiping a line between allergen and non-allergen runs instead of running the full validated changeover, on the assumption that a clean-looking surface means an allergen-free one.
Physically compensating for an awkward or high-frequency hand-pack motion instead of flagging it for task rotation or line rebalancing.
Overcorrection after being burned once: reweighing or re-swabbing every single unit going forward, including low-risk ones, which drags pace below the floor the role is also accountable for. The fix is targeted verification at the actual risk points โ sampling intervals, changeovers, fragility-class SKUs โ not blanket re-checking.
Worked example
Situation. A trail-mix packer on Line 3, Station 4 runs 12-count cartons (SKU TM-1206, "Net Wt. 1.5 oz / 42.5 g per bag") against a facility line-rate target of 900 cartons/hour (15/min). The checkweigher samples every 30th carton, weighing all 12 bags. 50 minutes into the shift the station has completed 640 cartons against a 750-carton target (15/min ร 50 min) โ a deficit of 110 cartons, running at 12.8 cartons/min, 85.3% of pace.
Naive read. Behind on rate with the deficit growing: the checkweigher flag on the current sample is "close enough" since the carton still rounds to 1.5 oz on the label, so keep scooping at the same dispenser setting and worry about the gap later.
Expert reasoning โ the checkweigher flag first. Sample #630 (the 21st skip-lot sample) reads 39.2, 39.5, 39.7, 41.0, 41.2, 41.5, 41.8, 42.0, 42.3, 42.6, 42.8, 43.0 g across the 12 bags. Sum = 496.6 g, average = 41.38 g โ below the 42.5 g label, which fails the lot-average requirement on its own. Separately, the per-bag MAV floor for this weight class (NIST Handbook 133's ~6% band for 1โ2 oz declarations) is 42.5 g ร 0.94 โ 39.95 g; three of the twelve bags (39.2, 39.5, 39.7 g) sit below that floor, a second, independent violation. Both point to the same root cause: the hand scoop is under-delivering, not a one-off light bag.
Expert reasoning โ quantify catching it now versus later. The shift's full production target is 7,200 cartons (900/hr ร 8 hr); a state weights-and-measures inspector auditing the day's lot against Handbook 133's sampling procedure would test the whole day's run, not just this sample. If the drift isn't caught, a failed-lot finding requires pulling and reworking the full day's production: at roughly $0.30/carton in relabel-and-repack labor, 7,200 ร $0.30 = $2,160, before any state civil penalty (amounts vary by jurisdiction) on top. Catching it now costs a station-level dispenser recalibration: about 4 minutes of this station's output (4 ร 15 = 60 cartons of delayed throughput) plus a 5-minute supervisor calibration check at $22/hr (โ$1.83), for roughly $8 total. That's a ratio of about 270x โ catching the drift now costs a fraction of a percent of what a failed lot costs later, and the pace gap it costs (an extra few minutes) is smaller than the 110-carton deficit already accrued from other causes this shift.
Resolution โ exception log, as filed:
Line 3, Station 4 โ Trail Mix 12-ct carton (SKU TM-1206), checkweigher sample #630 (21st skip-lot sample, shift 7/8/26).
12-bag weights (g): 39.2, 39.5, 39.7, 41.0, 41.2, 41.5, 41.8, 42.0, 42.3, 42.6, 42.8, 43.0 โ average 41.38 g vs. 42.5 g label average requirement; 3 of 12 bags below the 39.95 g individual MAV floor.
Held station, flagged scoop-dispenser calibration to line lead before resuming โ did not restart at the same setting.
Shift pace at time of flag: 640 of 750-carton 50-minute target (85.3% of rate, 110-carton deficit) โ logging as a calibration/staffing question, not widening the skip-lot interval to make up time.
Requesting scale/scoop recalibration check on Station 4 before resuming.
Logged by: [packer]. Line lead notified: [initials].
The naive read treated the checkweigher flag as noise to route around under pace pressure; the expert read treated it as a two-part legal-compliance failure with a quantifiable, order-of-magnitude cost difference between catching it at the station and catching it at a lot-level audit.
Going deeper
references/playbook.md โ load when running a net-weight/MAV check, selecting void-fill by fragility class, or executing an allergen changeover validation step by step.
references/red-flags.md โ load when reviewing a shift, a SKU, or a line's trend data for smell tests before a compliance failure, a damage claim, or an injury report.
references/vocabulary.md โ load when a term (MAV, cushion curve, changeover validation, HAL) needs a precise, misuse-aware definition.
Sources
National Institute of Standards and Technology, Handbook 133 โ Checking the Net Contents of Packaged Goods (current edition) โ the average-requirement and Maximum Allowable Variation concepts for sold-by-weight packages; exact MAV percentages by declared-weight class should be checked against the currently published table, treated here as planning-level bands.
U.S. Food and Drug Administration, Food Allergen Labeling and Consumer Protection Act (2004) and the FASTER Act (2021, effective 2023) โ the nine major food allergens requiring declaration and cross-contact control (milk, egg, fish, shellfish, tree nuts, peanuts, wheat, soybeans, sesame).
Sealed Air Corporation packaging-engineering literature and the ISTA (International Safe Transit Association) 3-Series test procedures โ cushion-curve methodology and fragility (G-rating) classification underlying void-fill selection.
American Conference of Governmental Industrial Hygienists (ACGIH), Threshold Limit Value for Hand Activity Level โ the HAL/peak-force framework for repetitive hand-intensive task risk, used as a stated heuristic for the reps-per-hour and force zones referenced here.
Benjamin Niebel and Andris Freivalds, Methods, Standards, and Work Design (McGraw-Hill) โ predetermined time systems and the speed-accuracy tradeoff in manual assembly/packing tasks underlying the line-rate-vs-defect-rate framing.
Bureau of Labor Statistics, Survey of Occupational Injuries and Illnesses, food and packaging manufacturing sector data โ basis for repetitive-motion/musculoskeletal disorders being a leading injury category in high-volume hand-packing.
No direct hand-packager practitioner has reviewed this file yet โ flag corrections or gaps via PR.
