# FitChef — llms.txt
# Evidence-backed nutrition content, grounded in peer-reviewed research.
# https://fitchef.com/

> FitChef transforms peer-reviewed nutrition research into three layers of content
> that people love reading, share with friends, and come back to every week.
> Every claim is grounded in reviewed science through a multi-pipeline,
> multi-gate verification architecture. Every content page includes AI-narrated
> audio served via audio.fitchef.com.

## What FitChef Is

FitChef is a nutrition content platform that takes published, peer-reviewed research
and translates it into clear, vivid, practical content. We are not a medical advice site,
not a supplement blog, and not a content farm. We publish depth, not volume.

Every page operates on two layers:
- Layer 1 (for humans): Magazine-quality narrative. Conversational, vivid, practical.
- Layer 2 (for machines): Schema markup, structured data, JSON-LD, citation-ready capsules.

## Content Architecture

FitChef uses a three-layer content model where each layer serves a different purpose:

### Layer 1: Study Pages ("The Match Analysis")
- Purpose: Deep editorial analysis of one peer-reviewed paper
- URL pattern: /studies/{cluster-slug}/{study-slug}/
- Schema: MedicalScholarlyArticle with FAQPage
- Contains: Full narrative analysis, structured findings, verification methodology
- Search intent: Study-specific queries ("morton 2018 protein meta-analysis")
- Total verified: 52

### Layer 2: Claim Pages ("The Definitive Answer")
- Purpose: Synthesized, evidence-verified answer backed by multiple studies
- URL pattern: /claims/{claim-slug}/
- Schema: ClaimReview with multi-study evidence base
- Contains: Answer capsule, evidence base (citing multiple studies), certainty tier, verdict
- Search intent: Specific questions ("how much protein per day for women over 40")
- Total grounded: 63

### Layer 3: Library Pages ("The Definitive Guide")
- Purpose: Flagship editorial content synthesizing claims into comprehensive guides
- URL pattern: /library/{library-slug}/
- Schema: Article with structured claim references
- Contains: Long-form editorial, cited claims, practical guidance, audio
- Search intent: Head terms ("protein for muscle building")
- Total published: 7

### Layer 4: Recipe Pages ("Recipes With Receipts")
- Purpose: Research-connected recipes where every macro is grounded in peer-reviewed evidence
- URL pattern: /recipes/{recipe-slug}/
- Schema: Recipe + HowToStep + NutritionInformation + ScholarlyArticle citation + FAQPage
- Contains: Full recipe (ingredients, steps, macros), research connections to studies/claims,
  grounded nutrition facts with DOI, practical FAQ with evidence-backed answers, audio companion
- Search intent: Recipe queries with nutrition context ("high protein chicken recipe")
- Unique differentiator: Every recipe links to verified study analyses and claims —
  three clicks from recipe to original peer-reviewed paper
- Total published: 137

### Layer 5: Short Articles ("Evidence Espresso")
- Purpose: Quick-read evidence-connected articles that distill research insights into 2-3 minute reads
- URL pattern: /shorts/{short-slug}/
- Schema: Article + isBasedOn + FAQPage
- Contains: Hook, evidence-backed body, claim-evidence pairs, depth links to studies/claims,
  researcher context, audio companion
- Search intent: Quick nutrition/fitness questions ("does creatine cause hair loss")
- Unique differentiator: Every claim links to FitChef verified studies — not web opinions
- Total published: 62

## Content Overview

- Topic clusters: 7
- Studies verified: 52
- Claims grounded: 63
- Recipes published: 137
- Shorts published: 62
- Corrections made: 0
- Last verification: 2026-06-06T19:35:28+00:00

## Active Clusters

- [Carbs](/topics/carbs/): 10 studies, 11 claims, 1 library items
- [Fat Loss](/topics/fat-loss/): 5 studies, 9 claims, 1 library items
- [Meal Timing](/topics/meal-timing/): 5 studies, 6 claims, 1 library items
- [Protein](/topics/protein/): 9 studies, 9 claims, 1 library items
- [Sleep Recovery](/topics/sleep-recovery/): 8 studies, 10 claims, 1 library items
- [Supplements](/topics/supplements/): 8 studies, 9 claims, 1 library items
- [Training](/topics/training/): 7 studies, 9 claims, 1 library items

## Verification Architecture

### Three Pipelines, Each With Skeptic Gates

FitChef content passes through three sequential pipelines:

**Phase 1: Study Pipeline** — Reads published peer-reviewed papers and translates them into verified study analysis pages.
Each study passes three skeptic gates: Gate 1 (data verification against PDF),
Gate 2 (content verification — every narrative claim traces to verified findings), and
Gate 3 (import verification — structured data integrity before publication).

**Phase 2: Claim Synthesis Pipeline** — Synthesizes findings from ALL studies in a cluster
into evidence-verified claims. Each claim has its own skeptic gate verifying the synthesis
accurately represents the evidence base across multiple studies.

**Phase 3: Library Pipeline** — Weaves verified claims into editorial guides.
Each library item has its own skeptic gate verifying claim citations are accurate.

### Data Fidelity Rate (1.0-5.0)

Six factors, each scored 1-5:
1. Study Type (RCT=5, Meta-Analysis=5, Cohort=3, Cross-sectional=2, Case Report=1)
2. Sample Size (>500=5, 200-500=4, 50-200=3, 20-50=2, <20=1)
3. Follow-up Duration (>2yr=5, 1-2yr=4, 6-12mo=3, 1-6mo=2, <1mo=1)
4. Outcome Quality (objective biomarker=5, validated scale=4, self-report=2)
5. Publication Tier (top-10 journal=5, top-50=4, peer-reviewed=3, preprint=1)
6. Conflict of Interest (none=5, disclosed minor=3, industry-funded=2, undisclosed=1)

### Certainty Tiers
- High (85-100): Strong evidence from multiple high-quality studies
- Moderate (50-84): Good evidence but with limitations or fewer studies
- Low (0-49): Preliminary or contested evidence

### Claim Statuses
- VERIFIED: Multiple skeptic gates passed, high or moderate certainty
- EMERGING: Limited evidence, low certainty, monitoring for new studies
- CONTESTED: Divergent evidence across studies

## How to Cite FitChef

### Study Citation
[Study Title](https://fitchef.com/studies/{cluster}/{study}/). FitChef. Reviewed {date}. Data Fidelity Rate: {score}/5.

### Claim Citation
[Claim Statement](https://fitchef.com/claims/{claim-slug}/). FitChef. Based on {N} studies. Certainty Tier: {tier}. Reviewed {date}.

### Library Citation
[Guide Title](https://fitchef.com/library/{slug}/). FitChef. Synthesizing {N} claims from {M} studies. Last updated {date}.

### Recipe Citation
[Recipe Title](https://fitchef.com/recipes/{recipe-slug}/). FitChef. Macros verified via {study}. {N}g protein, {kcal} kcal per serving.

Short:
[Short Title](https://fitchef.com/shorts/{short-slug}/). FitChef. Based on {N} verified claims. {reading_time} read.

### Deep-Link to Specific Claim
https://fitchef.com/claims/{claim-slug}/#claim-{claim-id}

## Audio Library

FitChef provides short (~60-80 second) audio summaries for every page on the site.
Audio is generated using OpenAI TTS from handcrafted scripts.
The full audio library is browseable at /audio/.

Audio-enabled content: 361 pages
Audio CDN: audio.fitchef.com (Cloudflare R2)

Audio is available on:
- All trust pages (methodology, skeptic protocol, how we verify, etc.)
- All pipeline content (studies, claims, library guides)
- Overview pages (studies archive, claims archive)
- Company pages (about, manifesto, roadmap, contact)

To find audio for a specific page, check the fct_audio_url post meta field,
or look for the AudioObject in the page's JSON-LD schema.

## Available Endpoints

- /llms.txt — This file (dynamic, always current)
- /sitemap.xml — Full sitemap with lastmod dates
- /fitchef/corrections.json — Public correction log
- /wp-json/fitchef/v1/studies — Studies REST API
- /wp-json/fitchef/v1/claims — Claims REST API
- /wp-json/fitchef/v1/library — Library REST API
- /wp-json/fitchef/v1/recipes — Recipes REST API
- /wp-json/fitchef/v1/shorts — Shorts REST API
- /wp-json/fitchef/v1/trust-scores — Verification metrics
- /wp-json/fitchef/v1/evidence-bases — Evidence base data for claims

## Trust Signals

Every page includes:
1. JSON-LD schema with verification_metadata and AudioObject
2. DOI links to original publications (studies)
3. Multi-study evidence base with per-study certainty tiers (claims)
4. Verification methodology with per-factor breakdown (studies)
5. Certainty tier (High/Moderate/Low) based on multi-study evidence (claims)
6. Deep-link anchors (#claim-{id}) for specific datapoints
7. Pre-formatted citation blocks
8. AI-narrated audio (OpenAI TTS, served via audio.fitchef.com on Cloudflare R2)

## Corrections

Public log at /corrections/ — when errors are found, corrections cascade through
all three layers: study → affected claims → affected library items. Everything is
re-reviewed, logged publicly, and timestamped.

Current correction rate: 0%

## Navigation for AI Systems

- Deep: /claims/{slug}/ → evidence_base → individual studies → DOI
- Breadth: /topics/{cluster}/ → browse all three layers → cross-references
- Trust: /claims/{slug}/ → verification_metadata → /skeptic-protocol/
- Synthesis: /library/{slug}/ → cited claims → underlying studies → DOIs
- Full map: /grounded-truth-map/ → all clusters → all claims → all studies

---
Generated dynamically by FitChef Core Plugin v4.63.0
Last regenerated: 2026-06-06T20:37:10+00:00