The Food Framework: Provenance First, and the Four Camps a Meal Can Sit In

Summary

The whole diet framework runs on one question you can ask without thinking — can I trace where this came from, and is this one ingredient or thirty? — because what holds the baseline down is the engineered-food matrix and liquid sugar rather than any single villain ingredient, and what lets it settle is returning to food whose origin you can name. Everything else here is that question worked out: the four camps a meal can sit in, the floor worth releasing first, why having enough food is a separate read from the quality of it, and the register the whole thing has to be held in — because anxiou

Why Foundational

Tier 0.5 because the components are well-established and the integration is Realised's synthesis. The recovery frame, the ultra-processed and liquid-sugar floor, protein at meals and blood-sugar stability as nervous-system substrate are each well supported independently; assembling them behind a single provenance question, with the camps as the reading and the held-lightly register as a specification, is the platform's own construction rather than a finding.

NOT Tier 1, because Tier 1 would imply the specific structure has been trialled. It has not. Nobody has run a comparison of a provenance-led heuristic against an ancestral one for real-world dietary change, and the camp ladder is authored.

NOT Tier 2, because the load-bearing planks are not weak. The controlled evidence under the engineered-matrix floor is as strong as anything in the diet pillar, and the sugary-drink evidence is close behind it.

Sub-claim tiers, which do not all sit at the entry's level:
• The floor — removing the engineered matrix and liquid sugar: Strong. Controlled feeding evidence on intake, plus a large and consistent literature on sugary drinks. The qualifier from food_quality about label versus properties travels with it.
• Protein present at meals, and blood-sugar stability as a nervous-system substrate: Strong. Both are conventional physiology rather than platform positions.
• The seed-oil demotion: Moderate, and it is a framing decision rather than a safety verdict. Independent evidence supports "marker rather than independent toxin"; it does not support "seed oils are good for you", and this entry claims only the first.
• Sucralose and emulsifiers: Emerging and contested, and deliberately excluded from anything that gates progress.
• The four camps and their thresholds: Moderate on direction, authored on magnitude. That home-prepared food tracks better diet quality is well replicated observationally; the specific shares at each rung are a rendering of a ladder and are flagged for recalibration.
• Provenance as the best available heuristic: Moderate, and reasoned rather than trialled. It is a better answer to the stated objections than the alternatives, which is a different and weaker claim than being demonstrated superior in use.

Practical takeaway

Ask the provenance question about defaults, not about plates. The framework is aimed at what you eat most of the time — the shop, the standing meals, the things that repeat. A single meal is never the unit of anything.

Start at the floor, in this order. Liquid sugar first, because it is the single biggest return for the least disruption. Then the engineered matrix, which mostly happens on its own once you are cooking. Protein present at meals, and enough food, run alongside both from the beginning rather than after them. The week-by-week sequence is in diet_practical_implementation.

Replace; never subtract. Every move in this framework has a destination. Sugary drink becomes water, sparkling water, tea or coffee. Refined oil becomes butter, ghee or olive oil. A ready meal becomes a meal you cooked. If a change leaves a hole rather than a substitution, it is the wrong change, and it will not hold.

Move up the camps one rung at a time. From ready-made to someone-else-cooked is a real improvement and deserves to be treated as one. From someone-else-cooked to your own kitchen is the largest single step available. From your kitchen to your kitchen-from-ingredients is the last one and the least urgent. Nobody starts at the top, and nothing here requires arriving there.

What settling looks like, in the body. Steadier energy across a day, without the mid-afternoon collapse. Hunger that arrives at meal times rather than continuously. Three-to-four-hour gaps between meals that stop feeling like an emergency. Waking through the night less often. These are the readings worth attending to, and they are all felt rather than measured.

Individual variation is expected inside the frame. More or fewer carbohydrates, dairy or no dairy, more red meat or more fish, three meals or two — all traditional human diets were variants of whole food without industrial processing, and the framework's non-negotiable is the absence of industrial interference rather than any particular ratio. Experiment freely inside that.

Evidence detail

Why This Entry Exists

Someone arrives having been told, for twenty years, that eating well is a performance: a macro ratio to find, a calorie budget to defend, a list of superfoods to buy, a list of poisons to avoid. They have usually tried several versions and found that each one worked for a while and then stopped, and that the trying itself became a low background hum of vigilance about food. The framework is not a better version of that. It is a way of putting it down.

Realised treats diet the way it treats every pillar: as restoration rather than achievement. The claim is not that there is a peak nutritional state you have never reached and should now climb toward. The claim is that the body already knows how to regulate appetite, hold a steady energy through a day, and stop eating when it has had enough — and that those systems work when they are not being interfered with. So the target is to eat what humans ate before metabolic disease was common, and the work is subtraction: remove the industrial interference, and the baseline the body already knows how to hold comes back on its own.

That reframe matters most because of what the modern food environment actually does to people. It leaves them overfed and undernourished at the same time — more calories available than at any point in history, alongside genuine shortfalls in fibre, micronutrients and omega-3. That combination is not something a body produces on its own; it takes engineered food to arrange. Which is why the recovery move is to leave the pathological food environment rather than get better at navigating it. Navigating it is a full-time cognitive job that never ends. Leaving it is a handful of defaults that stop needing decisions.

What bad advice this protects against:
• "Eating well means eating less." → It does not, and this is the correction that comes first every time. You replace a few industrial inputs with real food; you never go hungry, and you never count anything.
• "There are clean foods and dirty foods, and you should feel bad about the dirty ones." → There is no moral register here at all. Whole fruit is fine. Treats are fine. The framework applies to what you eat most of the time, never to the audit of a single plate.
• "Read every label and hunt the bad ingredients." → Label-hunting is precisely the anxious vigilance this framework exists to release, and it is also the less effective move: remove the engineered matrix and the worst additives leave with it, without any of them ever being named.
• "Seed oils are the poison; eliminate them and you are done." → The evidence has moved and the framework moved with it. Seed oils are best understood as a marker of the ultra-processed matrix rather than an independent toxin, and treating them as the enemy both overstates the case and distracts from the removal that actually carries the benefit.
• "Home-cooked means healthy." → It is a strong signal and it is not a guarantee, which is exactly why the second question exists.
• "You need to know a lot about nutrition to eat well." → The point of a provenance question is that it works without nutritional literacy. If a heuristic still requires you to be an expert in the grey zone, it has not removed the need for expertise; it has just moved it.

This entry OWNS: the provenance-led heuristic and its demotion of the older 100-Year Test; the four meal camps, the two questions that place a meal, and the rules about what a camp reading is and is not; the floor (the engineered matrix plus liquid sugar) as the framework's single highest-return removal; the seed-oil demotion as a framing decision; the separation of adequacy from quality; and the held-lightly register the whole framework must be delivered in.

This entry DEFERS: the measurable properties behind food quality and the honest argument about whether the "ultra-processed" label or the underlying properties carry the causal signal, to food_quality; the sugary-drink and satiety evidence to hidden_liquid_calories_and_satiety; the full seed-oil evidence file, including the precautionary case for minimising refined and repeatedly-fried industrial oils, to seed_oils_health_claims; sweetener and emulsifier evidence to sucralose_evidence_and_safety and food_additives_to_avoid; blood-sugar physiology to blood_sugar_regulation; gut and fibre biology to diet_gut_microbiome and dietary_fiber_diversity_and_microbiome_health; protein amounts to diet_protein_intake; and the week-by-week implementation order to diet_practical_implementation.

The Heuristic: Provenance First

The question, in full: can I trace where this came from, and is this one ingredient or thirty?

That is the load-bearing axis of the entire framework, and it was chosen deliberately over the alternatives. It catches the things a rule-based list misses — the flavoured yoghurt with a thirty-item ingredient list does not survive a traceability question even though yoghurt is an ancient food — and it is the same axis as the anti-ultra-processing one, arrived at from the direction a person can actually use in a shop. It asks nothing of your nutritional knowledge. It runs in about a second. And when it runs, it produces an answer you can act on rather than a judgement you have to feel something about.

The 100-Year Test is now a supporting check, and it used to be the headline. The older form of this framework led with it: if a food did not exist in 1925, approach it with suspicion. It is a good instinct and it is kept here, demoted, because it is right about the extremes and it fails in three specific places. It goes ambiguous exactly where most real decisions live — sourdough, whole-wheat pasta, an extracted oil versus the seed it came from — so the person still needs literacy in the grey band. It passes adulterated old foods, because age is not the active ingredient and a century-old recipe can still be a refined-flour-and-sugar delivery system. And "what we evolved to eat" has no single referent: humans ate as generalists, and the Inuit and the Kitavans were both metabolically healthy on diets that look nothing like each other. Provenance gets everything the 100-Year Test gets right and repairs those three failures. Where the two disagree, provenance decides.

The other lenses are worth naming so their limits are visible. Traditional or from-a-country is weak on its own, since plenty of traditional dishes are refined, sugared and fried. Home-cooked is a strong proxy for the absence of industrial processing and still not a guarantee of anything, because a home cook can fry in refined oil and bake with sugar. Imagine your ancestors eating this carries the "whose ancestors" problem and a nostalgia bias. Each is a supporting check. None of them ever carries the call alone.

The deeper reason this is the axis: the framework is trying to give you back a capacity to judge, rather than hand you a rulebook to comply with. That is the diet-pillar version of the platform's whole thesis. A rulebook has to be consulted, which means food stays a decision you are constantly making. A question that runs by itself stops being a decision at all — which is the state worth getting back to.

The Four Camps

Provenance is a question you ask in the moment. The camps are the same question made repeatable, so that a month of meals can be read at all without anyone counting anything.

Two questions place any meal. The first is who made this? The second is asked only when the answer to the first was "I did": did it come together from ingredients, or from packages? That is all. There is no third question, and neither of them is about how much you ate.

The four camps, in plain words:

1. It came ready-made. A factory made it, wherever you happened to eat it. Fast food sits here. So does a ready meal you heated at home, because heating something is not making it. So does the catered tray at an event, since nobody in the room cooked it.
2. Someone else cooked it. A restaurant, a takeaway that cooks to order, a friend's kitchen, a canteen with real cooking in it. Real food, real cooking, someone else's hands — and you did not choose what went in.
3. You cooked it. Your kitchen, your hands. This is the camp most people move into first, and it is the largest single change in the ladder.
4. You cooked it from ingredients. Your kitchen, and the meal came together from ingredients rather than from packages. This is the top of the ladder, and it is worth being precise about why: it is a fact about provenance, not a grade for the food. A meal built from ingredients is a meal whose entire composition passed through your hands, which is the practical definition of traceable.

What a camp reading is never. It is a fact about where food came from. It is never a judgement of the meal, never a quantity, never a calorie count, and never a reading of your body. There is no such thing as a bad meal in this system, and a camp is not a score you can do badly at.

Only the month's average is ever read. A single meal is one point among a month of them — on the order of forty readings, not one verdict — and no single meal is ever surfaced back to you as a result. That is the deliberate design: a system that reacted to individual meals would be a system that made you watch individual meals, which is the loop this framework exists to quiet.

A meal that cannot be read simply does not count. If there is no record of where it came from, it leaves the count entirely rather than being guessed into a camp. Nobody is penalised for a month that could not be read; a thin month reads as we cannot tell yet, and never as you failed. The same rule protects the reading itself, because a month assembled from guesses would not be a fact about anything.

The thresholds are provisional, and honestly so. The shares that mark each rung of the ladder are authored from the structure rather than read off a trial, and they are explicitly open to recalibration once there is real data to recalibrate against. The direction is well supported; the exact numbers are a first draft, and the entry says so rather than dressing them up.

The Floor

If only one thing gets done, it is this one.

The single highest-return dietary removal is the ultra-processed, hyperpalatable engineered-food matrix, together with liquid sugar. These are the foods running the appetitive loop — engineered so that the combination of sugar, fat, salt and texture overrides the signal that would otherwise tell you to stop, stripped of the fibre and water that would have slowed you down, and eaten fast enough that fullness arrives after the eating has finished. The strongest evidence available in the whole diet pillar sits under this claim, including controlled feeding work where the same people, given free access, eat substantially more on the engineered diet without noticing and without choosing to. The numbers, the mechanisms and the honest argument about whether the label "ultra-processed" or the underlying properties carry the causal signal all belong to food_quality, and that qualifier travels with this claim rather than being dropped from it.

Liquid sugar is the highest-impact single removal inside that floor. It bypasses satiety almost completely, moves blood glucose faster than anything else people consume, and is the largest source of added sugar in most diets. If a person changes one thing, it is this one. The evidence is in hidden_liquid_calories_and_satiety.

The worst additives leave as a class effect, and none of them is hunted by name. Remove the engineered matrix and refined seed oils, sucralose and emulsifiers go with it, because they are largely markers of that matrix rather than things you find in food you made. This is both the higher-return move and the non-neurotic one: hunting individual ingredients on labels is slow, it is the ingredient-policing this framework exists to prevent, and it is redundant once the matrix is gone.

On seed oils specifically, the framing has changed and this entry states the change plainly. Seed-oil elimination is no longer an independent non-negotiable and is folded into matrix removal. The reason is the evidence: independent cohort and meta-analytic work finds higher linoleic acid tracking lower inflammatory markers rather than higher, including in a very large biomarker cohort, and academic nutrition scientists reviewing the topic hold that the reputation comes substantially from conflating seed oils with the ultra-processed foods they mostly appear in. The symmetry check runs on the side now being agreed with, too: the loudest pro-seed-oil sources include soy-industry-funded work, which is why the framing rests on the independent evidence rather than on those. Cooking with butter, ghee, olive oil, tallow or lard remains the recommendation, with the rationale corrected — these are whole, traceable fats, which is a provenance argument, and not a claim that the alternatives are toxins. The full evidence file, including the genuinely defensible precautionary case for minimising refined and repeatedly-fried industrial oils, is owned by seed_oils_health_claims, and this entry does not restate or overrule it.

Sucralose and emulsifiers are named here only to be tiered honestly. Both carry real emerging signals — sucralose on glycaemic handling and the microbiome, emulsifiers on the gut barrier — and neither is settled. They sit at Emerging and contested, well below the floor, and they are removed as a byproduct rather than as a target. See sucralose_evidence_and_safety and food_additives_to_avoid.

Adequacy, Which Sits Beside Quality And Never Inside It

Two things are true at once and must never be merged into a single read.

A protein source present at meals. Protein is what makes the rest of this effortless: it produces satiety, blunts the glucose response of the meal it arrives with, protects muscle, and removes the willpower problem by removing the between-meal hunger that creates it. The twenty-to-thirty-gram figure people quote is best understood as a description of what a normal portion already weighs — three eggs, a piece of fish, a tin of tuna, a cup of yoghurt, a bowl of beans. It is about inclusion: is there a protein source at this meal? It is never a target to hit, a number to track, or a figure to fall short of. Amounts and sources live in diet_protein_intake.

And enough food. This is the half most food advice forgets. Eating too little is a failure of this framework rather than an advanced version of it — it drives the same appetitive loop the floor is meant to release, from the other direction. The framework never asks anyone to eat less, and any reading of it that produces restriction has been misread.

Adequacy is read separately from quality, deliberately. A month can be entirely home-cooked and still be too little food, and the system must be able to say so without that showing up as a quality problem. Fusing the two reads would produce exactly the metric that turns eating into a performance, which is the outcome this design refuses.

Where The Precision Lives

The camps are coarse on purpose, and it is worth understanding why rather than reading it as an oversight.

The camps carry only what is cheaply and honestly observable at every single meal: who made it, and whether it came together from ingredients or from packages. Everything finer than that — swapping to traditional fats, getting vegetables into the day, fermented foods, fibre diversity — is real work that genuinely matters, and it happens through the practices the platform serves rather than through grading meals. A traditional-fats swap is a specific practice with a specific place in the recovery sequence. Vegetables, fermented foods and fibre are their own practices with their own evidence (dietary_fiber_diversity_and_microbiome_health, fermented_foods, diet_gut_microbiome).

Putting composition into the camps would mean asking you to describe the contents of every meal, every day, forever — which is food-monitoring rebuilt under a new name. Putting composition into the practices means it arrives as something to do, once, when it is your turn to do it. The coarseness of the camps is what buys the precision everywhere else.

There is one honest gap in this design and it is named rather than hidden: someone whose meals are all home-cooked but who drinks sugary drinks alongside them will read well on camps while liquid sugar continues. That is a known limit of a provenance-only reading, and it is why liquid sugar is called out explicitly as the floor's highest-impact single removal rather than being left for the camps to catch.

Mechanism

Why diet sits inside a nervous-system recovery platform at all: blood-sugar stability is the substrate of nervous-system stability. Glucose volatility drives cortisol and adrenaline release; the crash on the other side of a spike presents subjectively as anxiety; chronic high insulin and low-grade inflammation both reach the brain. A nervous system running on a blood-sugar rollercoaster cannot be regulated by technique, which is why diet is treated as infrastructure for the mental pillar rather than as a separate concern. The physiology is blood_sugar_regulation; the gut arm is diet_gut_microbiome.

Why the engineered matrix is the target rather than any single ingredient. The matrix works by defeating a signal rather than by poisoning anything: energy density, softness, eating rate and engineered palatability combine to deliver more food than the satiety system can account for in the time available. That is a loop — the food produces the appetite that produces the food — and loops are what this platform is built to interrupt. Removing the matrix does not require willpower against the loop; it removes the input the loop runs on.

Two supporting levers, at their honest strength. Eating vegetables and protein before carbohydrates genuinely blunts a meal's glucose and insulin excursion, sometimes substantially, and the pooled evidence shows no long-term HbA1c benefit over standard advice — so it is a real, free tool for managing individual meals rather than the metabolic transformation it is sold as (food_order_glycemic_control). A short walk after a meal reliably reduces that meal's glucose rise, through a contraction-driven mechanism that works even when insulin signalling is impaired, with an honest magnitude in the region of ten to twenty percent and the same caution about long-term markers (post_meal_walks_glucose). Both are worth doing. Neither is load-bearing, and neither substitutes for the floor.

The Register

The anti-neurotic property of this framework is not a property of its content. The identical heuristic is liberating when it is held lightly and orthorexic when it is held anxiously, so the register is part of the specification rather than a hoped-for side effect.
• Applied to defaults, never to policing every bite. The framework describes what your normal looks like; it has no opinion about tonight.
• "Most of the time," never "always." There is no version of this that requires an unbroken record, and an unbroken record is not the goal state.
• No clean and dirty, no good and bad foods, no moralising. Whole fruit is fine. Treats are fine and normal. There is nothing to feel about a meal.
• Perfection is not the target and never was. The target is a baseline that holds without being managed.
• Anxious food-vigilance is itself the loop this exists to quiet. Someone who is monitoring every ingredient with a knot in their stomach has more of the thing recovery removes, not less — regardless of how accurate their monitoring is.

Understanding the framework is not the same as having recovered. Someone can classify foods perfectly and be doing it from exactly the anxious vigilance the framework is meant to release — in fact, anxious food-vigilance tends to produce excellent classification. Getting the framework is a knowledge state. The recovered state is effortless discernment with a quiet appetite, and the two are tracked separately for precisely this reason. Comprehension is never read as clearing.

Risks And Contraindications

• This framework is about the quality of what goes in, and never about quantity, calories, portion size, fasting, or body weight. Any rendering of it that produces restriction has been misread, and the misreading is the harm.
• Never frame food as clean, dirty, junk, good or bad, and never anchor identity to food or body. "I eat clean" and "I don't eat junk" are the disordered self-concept rather than the goal; the honest anchor is a capability, along the lines of "I cook with real ingredients."
• Never present a camp reading as a verdict on a person or a meal. It is a provenance fact about a month. A reading that arrives as judgement has broken the design.
• Never surface an unreadable month as a failure. It reads as we cannot tell yet, and the correct response is to say so.
• When the held-lightly register breaks, that is a signal rather than a compliance problem. Fear, guilt, fixation on labels, escalating restriction, or distress about a single meal are the off-pattern markers, and they route to the eating-disorder staging and support already in the platform (eating_disorder_body_image_diagnostic). The framework does not adjudicate that territory.
• Anyone with a history of disordered eating: keep this strictly to which inputs to swap, never to how much to eat, and consider whether professional support belongs alongside it. This content is safe for that history precisely because it is about removing harmful inputs rather than forbidding food — and it stops being safe the moment it is delivered as purity.
• Never let the floor become a reason to eat less. Removing the engineered matrix and liquid sugar means replacing them with real food, and someone who removes without replacing has made things worse rather than better.

Controversy

Nature: interpretive and partly self-corrective — the two live disagreements here are with the framework's own earlier positions.

Disagreement one: is the ancestral heuristic the right axis? The case for it is that it is memorable, it catches the extremes correctly, and it points at something real about the timing of industrial food and the arrival of metabolic disease. The case against it is the one that won: it is ambiguous in the grey zone where decisions actually live, it passes adulterated old foods, and it has no single referent, since human diets varied enormously and evolution optimises for reproduction rather than for a long healthy life. There is also a fair objection that any curated shortlist of "simple rules" is itself an evidence-dependent act which moves the uncertainty behind a curtain and appoints a curator. Realised's position: the frame survives — a heuristic that builds discernment beats a rulebook that demands compliance — and the axis changed to provenance, which is the honest response to the objections rather than a rebrand.

Disagreement two: are seed oils an independent harm? The elimination camp points to genuine novelty, unprecedented volume, a documented conflict of interest in the bodies that promoted them, real oxidation chemistry, and trials of swapping saturated fat for high-dose linoleic acid that failed to show a survival benefit. The dismissal camp points to controlled trials showing no rise in inflammatory markers and biomarker cohorts where higher linoleic acid tracks lower inflammation and lower mortality. Realised's position, and it is a reversal: seed oils are a marker of the ultra-processed matrix rather than an independent suppressor, so they are folded into matrix removal and never hunted at label level. The practice of cooking in traditional fats survives on a provenance rationale. The separate precautionary case for minimising refined and repeatedly-fried industrial oils remains defensible and is owned by seed_oils_health_claims, which is the entry to read for the full file. The bias check cuts both ways here and is stated rather than hidden: the loudest pro-seed-oil sources include soy-industry-funded work, so the framing rests on independent academic and cohort evidence; and this platform's own earlier position was the elimination one, which means the correction runs against our prior publication rather than toward it.

A third, quieter disagreement worth naming: whether "ultra-processed" is a coherent causal category at all, or whether the causal work is done by the underlying properties — energy density, eating rate, engineered palatability — that the label happens to correlate with. That argument is genuinely open, it does not change what to do, and it belongs to food_quality.

Cross-Pillar Connections

• Diet (diet_tier_05_philosophy): the original Tier 0.5 philosophy entry. This entry supersedes its headline heuristic — that entry still leads with the 100-Year Test and lists seed-oil elimination as an independent non-negotiable, and where the two conflict, this entry is current. Its material on the abundance paradox, the natural pattern and the nervous-system connection stands.
• Diet (diet_practical_implementation): the week-by-week order of operations. Read its Tier 1 list with the seed-oil correction from this entry applied.
• Diet (food_quality): the measurable properties behind all of this, and the honest limits of the "ultra-processed" label. The floor claim here rests on that entry's evidence.
• Diet (hidden_liquid_calories_and_satiety): why liquid sugar is the single highest-impact removal.
• Diet (seed_oils_health_claims): the full seed-oil file, including the precautionary minimisation case this entry does not restate and does not overrule.
• Diet (diet_protein_intake): amounts and sources, held as inclusion rather than as targets.
• Diet (dietary_fiber_diversity_and_microbiome_health, fermented_foods, diet_gut_microbiome): the composition work that lives in practices rather than in camps.
• Diet (food_order_glycemic_control, post_meal_walks_glucose): the two supporting glycaemic levers, at their honest magnitudes.
• Sleep (blood_sugar_regulation): the physiology under "diet is nervous-system infrastructure", including the three-in-the-morning waking pattern that often turns out to be nocturnal hypoglycaemia.
• Mental (eating_disorder_body_image_diagnostic): where the register's failure mode routes. When food talk turns to fear, guilt or restriction, this framework hands over rather than continuing.

Sources and Grounding

This entry is grounded by reference rather than by primary citation: it is the framework layer, and the evidence for each plank lives in the entry that owns it. That is deliberate, since duplicating figures across entries is how a corpus drifts out of agreement with itself.
• The floor (engineered matrix, intake, eating rate, hyperpalatability): food_quality, which carries the controlled inpatient feeding trial and the energy-density-versus-palatability decomposition with their funding notation and open verification flags.
• Liquid sugar: hidden_liquid_calories_and_satiety.
• Seed oils, both directions: seed_oils_health_claims, which carries the novelty and conflict-of-interest history, the recovered-data trials, and the biomarker-cohort findings on linoleic acid. The demotion recorded here rests on independent cohort and meta-analytic evidence finding higher linoleic acid tracking lower inflammatory markers, and on academic reviewers attributing the reputation substantially to conflation with ultra-processed foods; the symmetry note that the loudest pro-seed-oil sources include soy-industry-funded work is part of the record rather than an afterthought. Anyone re-verifying this demotion should run the receipt gate against that entry rather than against this summary.
• Sweeteners and emulsifiers: sucralose_evidence_and_safety and food_additives_to_avoid.
• Blood sugar, sequencing, post-meal movement: blood_sugar_regulation, food_order_glycemic_control, post_meal_walks_glucose.
• Framework provenance (internal): the consolidated diet framework specification of 2026 and the rulings of 16 August 2026, which set the two-question provenance scheme, redefined the top camp as cooked-from-ingredients, split social eating into someone-cooked and catered, and confirmed that composition tags feed practices rather than camps.

What would change our mind