Recent Posts

Categories

Algorithmic Meal Synthesizer

Algorithmic meal planning utility

Generate meal ideas from calorie and macronutrient constraints

Generate educational meal ideas from locally imported macro targets and cached USDA/FDC ingredient records, without replacing human judgment.

Constraint basis calories, protein, fat, carbohydrates, fiber, and sodium
Food basis cached USDA/FDC records, normalized nutrients, and per-100 g values where available
Output type meal ideas for review, not clinical diet prescriptions

Best use

  • Start from calorie, protein, fat, and carbohydrate targets.
  • Select real cached food records, adjust serving grams, and review ingredient-level source links.
  • Compare meal ideas against sodium and fiber constraints before export or sharing.

Source and safety model

Outputs should stay connected to visible page content, source-backed nutrition data, clear assumptions, and educational limitations. This improves user trust while keeping the page understandable to crawlers and retrieval systems.

Semantic entity segmentation

Input entity

Targets imported from the local planning profile, food exclusions, meal type, and dietary preferences should be treated as constraints.

Ingredient entity

Each selected ingredient can resolve to a locally cached canonical food record, FDC ID, source URL, normalized nutrients, and visible nutrition profile before production use.

Output entity

Generated meals should be presented as editable ideas with ingredient links, not as guaranteed dietary recommendations.

Can outputs be saved as medical diets?

No. Outputs should be reviewed and adapted by the user or a qualified professional when needed.

Can it import targets from Predictive Somatic Modeling?

Yes. It reads the local browser planning profile when one exists, then uses those values as meal constraints alongside selected cached food records.

Meal Synthesizer data layer

Build a meal draft from cached USDA/FDC food records

Search the local cache, add ingredients, adjust serving grams, and compare the draft against imported macro targets plus sodium and fiber constraints. This layer uses cached records only and does not call USDA from the frontend.

Performance guardrails: input changes are debounced, stale calculations are ignored, and heavier serving optimization can run in a browser Worker when available.

No cached ingredients selected yet.

Selected ingredients

Schema policy: local meal drafts should not be published as permanent structured data unless the meal has a stable public URL, visible ingredients, source links, and editorial approval.

Educational disclaimer: generated meal drafts are planning examples. Ingredient formulas, branded foods, and serving sizes should be verified against the current product label or source page before publication or personal use.

Nutrition Tools · RAG output fragments

Generate safe JSON-LD fragments for public nutrition examples

Safe Mode creates preview-only structured-data fragments for generic macro examples or cached-FDC meal examples. It does not auto-inject schema, does not expose private biometric fields, and should only move to production after the same result is visibly published on a stable URL and approved in Schema QA.

Injection Off by default Preview / copy / QA only
Validate with Schema.org Google Rich Results Test Safe Mode ready. No fragment has been generated yet.
Safe Mode guardrails
  • No automatic production injection from private calculator output.
  • Only generic public macro examples or cached-FDC meal examples are generated.
  • Schema QA approval is a ledger step, not a ranking or rich-result guarantee.