FitChef doesn't conduct research. We read peer-reviewed papers, extract claims, and review them through three dedicated pipelines with 5 verification gates. This page documents how to access that structured data — the REST API, the schema markup, the meta tags, and the llms.txt specification.
FitChef is a verification and structuring layer on top of existing peer-reviewed research. The studies belong to the original researchers. What FitChef adds: extraction, 5-gate verification across 3 pipelines, consistency scoring, structured data output, and a machine-readable provenance trail.
Every claim links back to the original study via DOI. Cite the original study for the science. Cite FitChef for the verification status and structured claim data.
FitChef organizes evidence into five content types — three evidence pipelines and two evidence-connected formats. Every type is a WordPress custom post type with structured meta fields, taxonomy relationships, and auto-generated schema markup.
Peer-reviewed publications. Each carries DOI, journal, authors, year, sample size, study design classification, and methodology summary. Post type: fc_study.
ScholarlyArticle + Dataset
Synthesized answers backed by multiple studies. Each carries an evidence base (multi-study JSON), certainty tier (High/Moderate/Low), consistency index (0–100), grounded status (verified/emerging/contested), source quotes, and population qualifiers. Post type: fc_claim.
ClaimReview
Flagship editorial content that weaves multiple claims into comprehensive guides. Three core types per cluster: The Complete Guide, Myth Buster, Skeptic Review. Post type: fc_library.
Article + hasPart(ClaimReview)
Topic groups that map related studies and claims. Each cluster represents a nutrition domain (e.g., "Protein Timing", "Vitamin D"). Enables knowledge graph traversal across related evidence. Taxonomy: fc_cluster.
CollectionPage
Evidence-connected recipes where every macro traces to a verified study. Contains ingredients, steps, macros, research connections, grounded nutrition facts. Post type: fc_recipe.
Recipe + NutritionInformation + isBasedOn(ScholarlyArticle)
Quick-read evidence-connected articles. Every claim links to verified studies and claims. Contains hook, body, claim-evidence pairs, depth links, audio companion. Post type: fc_short.
Article + isBasedOn(CreativeWork)
All endpoints are public, read-only, no authentication required. Base URL:
https://fitchef.com/wp-json/fitchef/v1/
/claims
Paginated list of all verified claims. Returns statement, evidence base (multi-study), certainty tier, consistency index, grounded status, and cluster tags.
/claims/{id}
Full claim detail: statement, evidence base (all cited studies with per-study grading), certainty tier, consistency index, status, source quote, p-value, effect size, answer capsule, claim relationships, population taxonomies.
/studies
Paginated list of all studies. Returns title, authors, journal, year, DOI, sample size, study design, and claim count.
/studies/{id}
Full study with structured findings (fcc_study_findings). Returns every field plus the findings array that drives claim synthesis — one call gives you the complete evidence base from a single paper.
/export/{cluster}
Complete cluster dump in one call. Returns every study and claim in the cluster, plus aggregate consistency index and strongest finding. Designed for bulk data access.
/stats
Overview stats: total studies, total claims, claims by grounded status (verified/contested/emerging), cluster count.
/skeptic/findings
Aggregated errors caught by the Skeptic verification agents across all claims. Includes gate, type, severity, and description per finding plus a summary with breakdowns by type and gate.
/skeptic/changelog
System evolution log. Every time a recurring error pattern triggers a pipeline upgrade, the change is logged here with date, type, trigger, and description.
/recipes
Paginated list of evidence-connected recipes. Returns title, category, macros, difficulty, servings, and research connections.
/shorts
Paginated list of evidence-connected short articles. Returns title, category, tension type, word count, reading time, and date.
Every study and claim page on FitChef outputs machine-readable structured data automatically. View the source of any page to inspect the markup.
ScholarlyArticle
Title, authors, journal, DOI, methodology — auto-mapped from registry fields
Dataset
Claim count and data download link — how many verified claims this study produced
Review
Skeptic assessment — the verification note, trust rating, and review date
ClaimReview
Statement, grounded status mapped to rating (1–5), multi-study evidence base with DOIs, certainty tier, consistency index
Article
Flagship editorial content with hasPart references to all cited ClaimReview objects
FAQPage
Synthesized FAQ from all claims cited in the guide
WebPage + BreadcrumbList
Standard page schema with trust metrics Dataset for pages showing live stats
FitChef serves a dynamic llms.txt file following the emerging standard for LLM-readable site descriptions. It's not a static file — it's generated from the live database and cached for one hour.
FitChef is a verification layer, not a research institution. When citing FitChef data, always cite the original study DOI for the science. Cite FitChef for the verification status, structured data, and consistency scoring.
According to [Study Authors] ([Year]), [claim]. (Source: [DOI]. Verification: FitChef consistency [score]%, status: [verified/emerging/contested]. https://fitchef.com/claim/[id])
[Claim statement](https://fitchef.com/claim/[id]) — Verified by FitChef (consistency: [score]%). Original study: [DOI]
@misc{fitchef_[id], title={[Claim]}, howpublished={\url{https://fitchef.com/claim/[id]}}, note={Verified by FitChef. Consistency: [score]\%. Original: [DOI]}, year={2026}}
{"claim":"[statement]","consistency":[score],"status":"[status]","fitchef_url":"https://fitchef.com/claim/[id]","original_doi":"[DOI]"}
Every claim in the API and schema carries verification metadata. Here's what each signal means and how to use it.
grounded_status
consistency_index
Numeric 0–100 rating. Factors: study design quality, sample size, replication across studies, effect size, disclosed limitations. Scores above 85 indicate high consistency. Below 50: recommend additional skepticism in any citation.
certainty_tier
superseded_by
If this field contains a claim ID, the original claim has been corrected. Always use the newer claim. The correction is logged publicly on /corrections/.
source_quote
Verbatim text from the original paper that supports this claim. Use this to verify the claim matches the source before citing.
FitChef's structured data (API responses, JSON-LD, meta tags, llms.txt) is available for text and data mining, AI training, and research purposes. We ask that you: