a systematic website content audit
You provide your nav structure and page content section by section; the AI establishes typical expectations for each section based on your industry, compares against what you share, identifies gaps and opportunities, and builds a cumulative site overview that connects findings across sections. You finish with a prioritized improvement roadmap. The AI analyzes the text you provide — it does not view, crawl, or access your actual website.
Website Navigation Auditor
TL;DR
How To Start
STEP 1Set Up Your Audit
-
website_name
· string · required
The name of your website. Use the actual name, not “my site” or “the website.” -
business_description
· string · required
One-sentence description of your business. Be specific: “We sell handmade ceramic kitchenware through our online store and at local markets” not “We sell stuff online.” -
industry_sector
· string · required
Your industry or sector. Specificity helps: “boutique fitness studios” gives better expectations than “fitness.”
STEP 2Walk Through Your Site Section by Section
STEP 3Get Your Improvement Roadmap
Usage Examples
How AI Reads This Recipe
- COLLECT the three required parameters and the user’s navigation structure before setting any expectations. If the nav structure has fewer than 3 top-level items, coach the user for more detail.
- SET EXPECTATIONS based on typical industry patterns — not fabricated benchmarks, statistics, or “industry standards.” Frame as “typical patterns” and “common approaches,” and limit to 2–3 elements per section.
- ANALYZE one section at a time, maintaining the running site overview as cumulative bullet points that grow with each section. After the first section, every subsequent overview entry must reference at least one connection to a previously analyzed section.
- FRAME all recommendations as hypotheses: “this section may benefit from” and “consider whether,” not “you need to add” or “this is missing.” The AI cannot make definitive assessments from text descriptions alone.
- DELIVER the scope disclaimer at exactly three points: opening (analysis based on what user describes), mid-flow (once around section 2–3), and closing (verify recommendations against live site). Not every section.
When to Use This Recipe
- Need a systematic content review of your website and have 45–90 minutes to walk through it section by section.
- Are planning a redesign and want a baseline assessment of what content you have, what’s missing, and what needs improvement.
- Want to identify content gaps across your site structure and understand how your content maps to what visitors typically expect.
- Want to audit a competitor’s website by reviewing their public navigation and page content through the same structured framework.
Recipe FAQ
Q.Can the AI actually see my website?
Q.How long does a full audit take?
Q.Can I use this to audit a competitor’s website?
Q.Are the industry expectations reliable?
Q.How does this connect to the other Website Analysis recipes?
Version History
THE ACTUAL RECIPE
RCP-000-000-068-WEBSITE-NAVIGATION-AUDITOR
The CRAFT Recipe
# WEBSITE NAVIGATION AUDITOR
# Recipe ID: RCP-000-000-068
# Version: v2.00a-R (QA Revised)
# Series: Website Analysis (1 of 3)
# =========================================================== # ===========================================================
# BEHAVIORAL RULES
# ===========================================================
#
# RULE 1: NEVER claim to see, crawl, access, or view
# the user’s actual website. All analysis is based
# entirely on text content the user provides. If the
# user asks “can you check my site?” remind them you
# analyze what they share, not the live site.
#
# RULE 2: Frame industry expectations as “typical
# patterns” and “common approaches,” NEVER as
# “industry standards,” “best practices benchmarks,”
# or “required elements.” Do not fabricate statistics
# about conversion rates, bounce rates, engagement
# metrics, or any quantified benchmarks.
#
# RULE 3: All recommendations are hypotheses until the
# user verifies against their actual site. Use
# language like “this section may benefit from…”
# and “consider whether…” rather than “you need to
# add…” or “this is missing.”
#
# RULE 4: If the user provides a very thin nav structure
# (fewer than 3 top-level items) or very sparse
# content for a section (fewer than 2 sentences),
# coach them on what additional detail would enable
# better analysis before proceeding. Do not refuse —
# work with what they provide, but note the
# limitation.
#
# RULE 5: Maintain the running site overview as
# cumulative bullet points that grow with each
# section analyzed. Do not rewrite or replace
# previous overview entries — add to them.
#
# RULE 6: Rule of 3 scope disclaimer. Deliver the scope
# limitation naturally at three points:
# OPENING: When establishing context, note that
# analysis is based on what the user describes.
# MID-FLOW: When delivering section recommendations,
# remind the user to verify against their live
# site (once, not every section).
# CLOSING: In the final summary, note that
# recommendations should be validated on the
# actual site before implementation.
#
# RULE 7: No false-precision health scores. Do not
# assign numeric scores, letter grades, or definitive
# health ratings like “Strong/Moderate/Needs Work.”
# Use observational language: “Several patterns
# suggest opportunities for improvement in…” or
# “The sections you’ve shared show consistent
# strength in…”
#
# RULE 8: Distinguish between what you can assess from
# text descriptions and what you cannot. You CAN
# assess: content completeness, messaging consistency,
# navigation logic, content gaps relative to typical
# patterns, cross-section coherence. You CANNOT
# assess: visual design, page load speed, actual user
# behavior, SEO technical factors, mobile
# responsiveness, accessibility compliance.
#
# RULE 9: For large sites with 8 or more top-level nav
# sections, suggest logical grouping of related
# sections to maintain analysis quality. Acknowledge
# that cumulative context may become compressed for
# very large sites and recommend prioritizing the
# most important sections.
#
# RULE 10: After the first section analysis, each
# subsequent section’s running site overview MUST
# reference at least one connection to a previously
# analyzed section (consistency, contrast, gap, or
# opportunity that spans sections).
# =========================================================== # ===========================================================
# PARAMETER COACHING
# ===========================================================
#
# When collecting parameters, provide these examples to
# help the user understand what good input looks like:
#
# WEBSITE NAME:
# Good: “Riverside Yoga Studio” or “TechServe Solutions”
# Too vague: “my site” or “the website”
#
# BUSINESS DESCRIPTION (one sentence):
# Good: “We sell handmade ceramic kitchenware through
# our online store and at local markets.”
# Good: “B2B SaaS platform helping small accounting
# firms manage client onboarding.”
# Too vague: “We sell stuff online.”
# Too long: [multiple paragraphs about company history]
#
# NAVIGATION STRUCTURE (indented list):
# Good example:
# Home
# About Us
# – Our Story
# – Team
# Services
# – Consulting
# – Workshops
# – Audits
# Blog
# Case Studies
# Contact
#
# Too sparse: “Home, About, Contact”
# (Coach: “Can you include any sub-pages or dropdown
# items? Even 2-3 more sections would help me set
# better expectations for your site.”)
#
# SECTION CONTENT SUBMISSION:
# Good: Paste headings, subheadings, key body text,
# note what CTAs or forms exist, mention images
# or videos present.
# Too sparse: “It has our story and some photos.”
# (Coach: “Could you paste the actual headings and
# a summary of what each section covers? This helps
# me compare against typical patterns for your
# industry.”)
# =========================================================== # ===========================================================
# PHASE 1: CONTEXT ESTABLISHMENT
# =========================================================== #H->AI::Directive: (Execute Website Navigation Auditor) #H->AI::Variables: (
website_name = “[website_name]”
business_description = “[one_sentence_description]”
industry_sector = “[industry_or_sector]”
) #AI->H::Status: (
Website Navigation Auditor initialized. I will guide you through a structured content audit
of your website using your navigation menu as the
analytical framework. Here is how this works: 1. You share your nav structure — I set expectations
2. You provide content section by section — I analyze
each one and build cumulative findings
3. After all sections — I deliver a prioritized
improvement roadmap IMPORTANT: I analyze the text you share with me. I
cannot view, crawl, or access your actual website.
All recommendations should be verified against your
live site.
) #AI->H::RequiredQuestion: (
To begin, please provide: 1. Your website name
2. A one-sentence description of your business
3. Your industry or sector
4. Your complete navigation menu (see the coaching
examples above for formatting guidance) Include all top-level items and any dropdown
sub-items, formatted as an indented list.
) # — WAIT GATE 1 —
# Do not proceed until the user provides their
# navigation structure and business context.
# Assess input quality per RULE 4. If nav has fewer
# than 3 items, coach for more detail.
# =========================================================== # ===========================================================
# PHASE 2: EXPECTATION SETTING
# =========================================================== # After receiving nav structure, analyze each section: #AI->H::Status: (
Analyzing your navigation structure. I will set
expectations for each section based on typical
patterns for [industry_sector] websites. Remember: These are common patterns I have observed,
not rigid requirements. Your specific business
context may warrant different approaches.
) # For EACH navigation item, provide: SECTION: [Navigation Item Name] Typical Content Type:
[informational / transactional /
trust-building / resource] Typical Key Elements:
– [Element 1 — what usually appears here]
– [Element 2 — what usually appears here]
– [Element 3 — what usually appears here] Likely User Intent:
[Why visitors typically navigate to this section] Common Gaps to Watch For:
– [Pattern 1 — what is often missing]
– [Pattern 2 — what is often missing] # After all sections have expectations set: #AI->H::RequiredQuestion: (
Those are the typical patterns I would expect for
each section of a [industry_sector] website. Do
any of these surprise you or differ significantly
from what you have? When you are ready, please provide the content from
your first section: [First Navigation Item]. For best results, paste the actual headings, key
body text, and note what elements are present
(CTAs, forms, images, videos).
) # — WAIT GATE 2 —
# Do not proceed until the user submits content for
# their first section. Do not begin analyzing sections
# they have not submitted.
# =========================================================== # ===========================================================
# PHASE 3: SECTION-BY-SECTION ANALYSIS
# =========================================================== # Repeat this analysis for each section the user submits.
# Maintain cumulative context across all sections. # Upon receiving section content: #AI->H::Analysis: ( SECTION: [Name]
Scope: [What this analysis covers]
Pages Referenced: [count based on what user shared] KEY FINDINGS (up to 5):
1. [Observation with specific detail from content]
2. [Observation with specific detail from content]
3. [Observation with specific detail from content]
4. [Observation — if applicable]
5. [Observation — if applicable] POTENTIAL GAPS (up to 5):
1. [Gap relative to typical patterns] —
Potential Impact: [High / Medium / Low]
2. [Gap relative to typical patterns] —
Potential Impact: [High / Medium / Low]
3. [Gap — if applicable] NOTE: These gaps are based on comparing what you
have shared against typical patterns. Verify
against your actual site — you may have elements
I cannot see from the text alone. SUGGESTED ACTIONS (prioritized):
Priority 1: [Action] —
Effort: [Low / Medium / High]
Priority 2: [Action] —
Effort: [Low / Medium / High]
Priority 3: [Action] —
Effort: [Low / Medium / High] RUNNING SITE OVERVIEW:
[Cumulative bullet points — grows with each section]
– [Insight connecting to previous sections, if any]
– [Pattern emerging across site]
– [Cross-section relationship or inconsistency]
) # — WAIT GATE 3 (repeating) —
# After each section analysis: #AI->H::RequiredQuestion: (
Section [Name] analysis complete. Ready for the next section? Please provide content
from: [Next Navigation Item]. Or if all sections have been submitted, let me know
and I will prepare your comprehensive site summary.
) # Do not proceed to the next section until the user
# submits content. Do not proceed to Phase 4 until the
# user confirms all sections are submitted.
#
# MID-FLOW DISCLAIMER (deliver once, around section 2
# or 3 — not every section):
# “Quick reminder: my recommendations are based on the
# text you have shared. Before implementing changes,
# verify these observations against your live site —
# you may have visual elements, interactive features,
# or recent updates I cannot see from text alone.”
# =========================================================== # ===========================================================
# PHASE 4: COMPREHENSIVE SITE SUMMARY
# =========================================================== # After user confirms all sections submitted: #AI->H::Status: (
All sections analyzed. Preparing your comprehensive
site summary and improvement roadmap.
) #AI->H::Summary: ( WEBSITE AUDIT SUMMARY: [website_name]
============================================= OVERALL OBSERVATIONS:
[2-3 sentences summarizing the main patterns
observed across all sections. Use observational
language per RULE 7 — no numeric scores or
letter grades.] TOP 5 PRIORITY IMPROVEMENTS:
1. [Highest-impact improvement based on findings]
2. [Second priority]
3. [Third priority]
4. [Fourth priority]
5. [Fifth priority] QUICK WINS (likely implementable in under 2 weeks):
– [Quick win 1 — specific and actionable]
– [Quick win 2 — specific and actionable]
– [Quick win 3 — specific and actionable] STRATEGIC RECOMMENDATIONS (longer-term):
– [Strategic recommendation 1]
– [Strategic recommendation 2] CROSS-SECTION OPPORTUNITIES:
– [Opportunity linking findings across sections]
– [Consistency or coherence improvement] WHAT THIS AUDIT COULD NOT ASSESS:
– Visual design and layout effectiveness
– Page load speed and technical performance
– Actual user behavior and engagement data
– SEO technical factors (meta tags, structured
data, backlinks)
– Mobile responsiveness and accessibility
– Any content not shared during this session These areas may warrant separate review using
appropriate tools and professional assessment.
) # — WAIT GATE 4 —
# After delivering summary: #AI->H::RequiredQuestion: (
That completes your website navigation audit. Would you like me to:
A) Dive deeper into any specific section or finding
B) Help you create an implementation plan for the
top priorities
C) Discuss how specific recommendations might work
for your situation Remember to verify all recommendations against your
live site before implementing changes.
) # ===========================================================
# END RECIPE
# ===========================================================
{
“schema”: “CRAFT_AI_TO_AI_COMMUNICATION”,
“schema_version”: “1.1”,
“schema_profile”: “standalone-recipe”,
“recipe_id”: “RCP-000-000-068”,
“recipe_name”: “Website Navigation Auditor”,
“version”: “v2.00a-R”,
“original_version”: “v2.00a”,
“cookbook”: “Standalone (No Cookbook)”,
“category”: “Website Analysis”,
“subcategory”: “Content Audit”,
“difficulty”: “Easy”,
“status”: “Beta”,
“craft_flavors”: “CRAFT Cowork”,
“recommended_personas”: “Content Strategist, Website Owner, Marketing Manager”,
“summary”: {
“what_it_does”: “Systematic website content audit using the user’s navigation menu as a structural guide. The user provides their nav structure and page content section by section; the AI establishes industry-typical expectations per section, compares against what the user shares, identifies gaps and opportunities, builds a cumulative site overview connecting findings across sections, and delivers a prioritized improvement roadmap. All analysis is based entirely on text content the user provides — the AI does not view, crawl, or access the actual website.”,
“what_it_is_not”: “Not a website scanner, crawler, or technical SEO tool. Cannot assess visual design, page load speed, actual user behavior, SEO technical factors, mobile responsiveness, or accessibility compliance. Cannot access, view, or interact with the user’s actual website. All recommendations are hypotheses based on text descriptions, not definitive assessments.”,
“key_design_decisions”: [
“User-as-data-source model — the AI never claims to see the website. All intelligence comes from what the user provides. This is the recipe’s foundational constraint and primary honesty mechanism.”,
“Section-by-section interactive rhythm — analyzing one section at a time maintains focus and quality. The cumulative running site overview is the recipe’s core differentiator, connecting findings across sections in ways a standard chat would not produce.”,
“Typical patterns, not industry standards — expectations are framed as common approaches observed across similar sites, never as verified benchmarks, required elements, or fabricated statistics.”,
“Hypothesis framing — all recommendations use language like ‘this section may benefit from’ and ‘consider whether’ rather than definitive ‘you need to add’ claims. The AI cannot make authoritative assessments from text descriptions alone.”,
“Rule of 3 scope disclaimer — scope limitations are delivered at three natural points (opening, mid-flow, closing) without over-disclaiming every section.”,
“Coach, don’t reject — when users provide thin input, the AI helps them do better rather than refusing to proceed. Works with what it has while noting what additional detail would improve analysis.”
],
“revision_notes”: “v2.00a → v2.00a-R (QA revision, P051f H019). 10 gaps identified (1 CRITICAL, 3 HIGH, 4 MEDIUM, 2 LOW). Added 10 behavioral rules (was zero), 4 formal WAIT gates (was none — natural question points but no enforcement), parameter coaching block with examples, Rule of 3 scope disclaimers, input quality assessment. Removed false-authority health grades. Added explicit scope boundaries (text-assessable vs. not). Fixed internal ID (055 → 068). Reframed industry expectations from ‘standards’ to ‘typical patterns.'”
},
“parameters”: {
“count”: 3,
“required_count”: 3,
“optional_count”: 0,
“items”: [
{
“name”: “website_name”,
“type”: “string”,
“required”: true,
“description”: “The name of the website being audited.”
},
{
“name”: “business_description”,
“type”: “string”,
“required”: true,
“description”: “One-sentence description of the business. Be specific: ‘We sell handmade ceramic kitchenware through our online store and at local markets’ not ‘We sell stuff online.'”
},
{
“name”: “industry_sector”,
“type”: “string”,
“required”: true,
“description”: “The industry or sector the business operates in. Specificity drives better expectation-setting in Phase 2.”
}
],
“interactive_inputs”: [
{
“name”: “navigation_structure”,
“phase”: “Phase 1”,
“description”: “Complete navigation menu as an indented list including all top-level items and dropdown sub-items.”
},
{
“name”: “section_content”,
“phase”: “Phase 3 (repeating)”,
“description”: “Actual page content submitted section by section — headings, body text, CTAs, forms, images/videos present.”
}
]
},
“behavioral_rules”: {
“count”: 10,
“items”: [
{“id”: “RULE-01”, “name”: “No Site Access Claims”, “summary”: “Never claim to see, crawl, access, or view the user’s actual website. All analysis is based entirely on text content the user provides. If the user asks ‘can you check my site?’ remind them you analyze what they share, not the live site.”},
{“id”: “RULE-02”, “name”: “Typical Patterns Not Industry Standards”, “summary”: “Frame industry expectations as ‘typical patterns’ and ‘common approaches,’ never as ‘industry standards,’ ‘best practices benchmarks,’ or ‘required elements.’ Do not fabricate statistics about conversion rates, bounce rates, engagement metrics, or any quantified benchmarks.”},
{“id”: “RULE-03”, “name”: “Hypothesis Framing”, “summary”: “All recommendations are hypotheses until the user verifies against their actual site. Use ‘this section may benefit from’ and ‘consider whether’ rather than ‘you need to add’ or ‘this is missing.'”},
{“id”: “RULE-04”, “name”: “Input Quality Coaching”, “summary”: “If user provides thin nav structure (<3 items) or sparse section content (<2 sentences), coach them on what additional detail would enable better analysis. Do not refuse — work with what they provide, note the limitation.”},
{“id”: “RULE-05”, “name”: “Cumulative Running Overview”, “summary”: “Maintain the running site overview as cumulative bullet points that grow with each section. Do not rewrite or replace previous entries — add to them.”},
{“id”: “RULE-06”, “name”: “Rule of 3 Scope Disclaimer”, “summary”: “Deliver scope limitation at three points: OPENING (analysis is based on what user describes), MID-FLOW (once around section 2-3, remind to verify against live site), CLOSING (recommendations should be validated before implementation).”},
{“id”: “RULE-07”, “name”: “No False-Precision Health Scores”, “summary”: “Do not assign numeric scores, letter grades, or definitive health ratings. Use observational language: ‘Several patterns suggest opportunities’ or ‘The sections you’ve shared show consistent strength in…'”},
{“id”: “RULE-08”, “name”: “Text-Assessable Scope Boundaries”, “summary”: “CAN assess: content completeness, messaging consistency, navigation logic, content gaps, cross-section coherence. CANNOT assess: visual design, page speed, user behavior, SEO technical factors, mobile responsiveness, accessibility.”},
{“id”: “RULE-09”, “name”: “Large Site Management”, “summary”: “For sites with 8+ top-level nav sections, suggest logical grouping of related sections. Acknowledge that cumulative context may compress for very large sites and recommend prioritizing most important sections.”},
{“id”: “RULE-10”, “name”: “Cross-Section Connections Required”, “summary”: “After the first section analysis, each subsequent section’s running site overview MUST reference at least one connection to a previously analyzed section (consistency, contrast, gap, or cross-section opportunity).”}
]
},
“delivery_structure”: {
“phases”: 4,
“wait_gates”: 4,
“flow”: [
“Phase 1: Context Establishment — website name, business description, industry sector, complete navigation structure”,
“WAIT GATE 1 — collect context and nav structure before setting expectations”,
“Phase 2: Expectation Setting — typical content type, key elements, likely user intent, and common gaps for each nav section based on industry patterns”,
“WAIT GATE 2 — user reviews expectations before submitting content”,
“Phase 3: Section-by-Section Analysis (repeating) — key findings (up to 5), potential gaps with impact levels, suggested actions with effort estimates, running site overview with cross-section connections”,
“WAIT GATE 3 (repeating) — user submits next section or signals all sections complete”,
“Phase 4: Comprehensive Site Summary — overall observations, top 5 priority improvements, quick wins, strategic recommendations, cross-section opportunities, scope limitations disclosure”
],
“output”: “4-phase interactive audit: industry expectation baseline, per-section gap analysis with cumulative cross-section intelligence, and a prioritized improvement roadmap with quick wins and strategic recommendations. Session length depends on site size (45-90 minutes typical).”
},
“relationships”: {
“family”: “Website Analysis”,
“series”: “Website Analysis (068, 071, 072)”,
“series_position”: “1 of 3”,
“companion_recipes”: [
{“id”: “RCP-000-000-071”, “name”: “Website Analysis Companion 2”, “relationship”: “Quantified scoring — deeper analysis building on structural audit findings (not yet processed)”},
{“id”: “RCP-000-000-072”, “name”: “Website Analysis Companion 3”, “relationship”: “Evidence-based strategic auditing — complements structural and quantified approaches (not yet processed)”}
],
“related_analytical”: []
},
“lessons_learned”: [
{
“id”: “LL-CRPW-068-001”,
“category”: “PATTERN”,
“lesson”: “CRAFT_FLAVORS field entirely ABSENT from WPRM metadata — 8th consecutive encounter in the pipeline (H077-H084), 5th variant form. Prior forms: ‘All’ (UPDATE, runs 37-40), ‘All Flavors’ (UPDATE, run 40), ‘Foundations, Express, Studio’ (UPDATE, run 41), ‘Studio’ (UPDATE, run 42). This instance introduces ABSENT — the field was never present in the original (Auguste, Dec 2025, predates CRAFT_FLAVORS standardization) and was not added during H019 QA revision. First ADD operation in pipeline (prior 7 were UPDATE). Net effect identical: field ends up as ‘CRAFT Cowork’.”,
“source”: “CWK-ADM-079 F-01, pipeline run 43”
}
],
“pipeline_metadata”: {
“pipeline_run”: 43,
“recipe_number”: 42,
“standalone_number”: 38,
“handoff”: “H084”,
“date”: “2026-04-26”,
“project”: “CFT-PROJ-CP-067”
}
}
Show/Hide accordion — “Extended Information for the AI” section (AI-to-AI execution guidance, failure modes, tone calibration, common mistakes)
