@growthub/cli 0.14.4 → 0.14.6

package/assets/worker-kits/growthub-geo-seo-v1/skills.md

@@ -1,359 +0,0 @@
-# GEO SEO Operator — Master Skill Doc
-
-**Source of truth for methodology. Read this file completely before beginning any task.**
-
----
-
-## QUICK REFERENCE TABLE
-
-| Resource | Path |
-|---|---|
-| Agent operating law | `workers/geo-seo-operator/CLAUDE.md` |
-| Brand kit template | `brands/_template/brand-kit.md` |
-| Growthub example brand kit | `brands/growthub/brand-kit.md` |
-| Output contract | `output-standards.md` |
-| Runtime assumptions | `runtime-assumptions.md` |
-| Fork integration notes | `docs/geo-seo-fork-integration.md` |
-| Subagent dispatch | `docs/subagent-dispatch.md` |
-| Scoring methodology | `docs/scoring-methodology.md` |
-| PDF report layer | `docs/pdf-report-layer.md` |
-| GEO audit brief | `templates/geo-audit-brief.md` |
-| Citability analysis | `templates/citability-analysis.md` |
-| Crawler access report | `templates/crawler-access-report.md` |
-| Brand visibility report | `templates/brand-visibility-report.md` |
-| GEO score summary | `templates/geo-score-summary.md` |
-| Content analysis | `templates/content-analysis.md` |
-| Schema validation | `templates/schema-validation.md` |
-| Technical foundations | `templates/technical-foundations.md` |
-| llms.txt plan | `templates/llmstxt-plan.md` |
-| Remediation roadmap | `templates/remediation-roadmap.md` |
-| Client proposal | `templates/client-proposal.md` |
-| Sample audit brief | `examples/geo-audit-sample.md` |
-| Sample citability | `examples/citability-sample.md` |
-| Sample PDF data | `examples/pdf-report-sample.md` |
-| Sample proposal | `examples/prospect-proposal-sample.md` |
-
----
-
-## STEP 0 — BEFORE ANY TASK, ANSWER THESE QUESTIONS
-
-Before producing anything, confirm:
-
-1. Which client or brand is this for?
-2. What is the target URL or domain?
-3. What is the audit scope: quick / citability-only / full / specific command?
-4. What delivery format is required: Markdown only / PDF / both?
-5. Is the geo-seo-claude fork available locally?
-6. What execution mode: local-fork / agent-only / hybrid?
-
-If any of these are unknown after the 3-question gate in CLAUDE.md, stop and ask.
-
----
-
-## STEP 1 — LOAD THE BRAND KIT
-
-Read `brands/<client-slug>/brand-kit.md` if it exists. Otherwise start from `brands/_template/brand-kit.md`.
-
-Extract:
-- client identity (name, slug, industry)
-- target URL and competitor URLs
-- audit scope and delivery format preference
-- messaging tone and guardrails
-- agency context (prospect stage, retainer range)
-- existing deliverables log
-
-The brand kit drives all output naming, tone calibration, and proposal pricing context.
-
----
-
-## STEP 2 — CHECK THE WORKING SUBSTRATE
-
-If the user has a local geo-seo-claude fork, inspect it before planning anything.
-
-### Source-of-truth file order in the fork
-
-1. `README.md`
-2. `geo/` — main skill entry point and command routing
-3. `skills/` — 14 sub-skill definition files
-4. `agents/` — 5 subagent definition files
-5. `scripts/fetch_page.py` — fetches and parses pages via Playwright
-6. `scripts/citability_scorer.py` — runs the 5-metric citability algorithm
-7. `scripts/brand_scanner.py` — scans platforms for brand mentions
-8. `scripts/generate_pdf_report.py` — produces branded PDF via ReportLab
-9. `scripts/llmstxt_generator.py` — generates llms.txt and llms-full.txt
-10. `scripts/crm_dashboard.py` — Flask-based CRM dashboard
-11. `schema/` — 6 JSON-LD templates for structured data recommendations
-
-### What to verify in the fork
-
-- Which of the 14 commands are implemented and callable
-- Whether Playwright is installed and chromium browser is available
-- Whether `requirements.txt` dependencies are installed
-- What the Python version is (`python3 --version`)
-- Whether `schema/` contains JSON-LD templates for all common types
-- Whether `agents/` has definitions for all 5 subagents
-
-If the fork cannot be inspected, use the frozen assumptions in `runtime-assumptions.md` and label outputs `assumption-based`.
-
----
-
-## STEP 3 — COMMAND SELECTION LOGIC
-
-Select the narrowest command that satisfies the real job.
-
-| Command | Primary Use | Phase 2 Subagents? | PDF Output? |
-|---|---|---|---|
-| `/geo audit` | Full GEO + SEO audit, all components | Yes | Optional |
-| `/geo citability` | Citability score only, fast | No | No |
-| `/geo crawlers` | AI crawler permission check | No | No |
-| `/geo brands` | Brand mention and authority scan | No | No |
-| `/geo report` | Structured Markdown report from existing data | No | No |
-| `/geo report-pdf` | Branded PDF report via ReportLab | No | Yes |
-| `/geo content` | E-E-A-T and content quality | Partial | No |
-| `/geo schema` | Structured data validation | No | No |
-| `/geo technical` | Server headers and technical health | No | No |
-| `/geo llmstxt` | Generate llms.txt plan or file | No | No |
-| `/geo quick` | 60-second AI visibility snapshot | No | No |
-| `/geo proposal` | Client proposal with pricing | No | No |
-| `/geo prospect` | Prospect qualification scan | Partial | No |
-| `/geo compare` | Side-by-side GEO comparison of 2+ URLs | Yes | Optional |
-
-Default selection rules:
-- "Full audit" → `/geo audit`
-- "Quick look" or "first pass" → `/geo quick`
-- "Check if crawlers can access" → `/geo crawlers`
-- "Can AI cite this page?" → `/geo citability`
-- "Build a proposal for this prospect" → `/geo proposal`
-- "Compare to competitors" → `/geo compare`
-
----
-
-## STEP 4 — PHASE 1: FETCH AND PARSE LOGIC
-
-`fetch_page.py` uses Playwright (chromium) to load the target URL dynamically and extract all signals needed for Phase 2.
-
-### What to extract
-
-| Signal | Source | Notes |
-|---|---|---|
-| `robots.txt` | `https://domain.com/robots.txt` | Parse Disallow/Allow rules for each of the 14 AI crawler user-agents |
-| `llms.txt` | `https://domain.com/llms.txt` | Exists / missing / malformed — record file contents if present |
-| `llms-full.txt` | `https://domain.com/llms-full.txt` | Exists / missing |
-| `sitemap.xml` | `https://domain.com/sitemap.xml` | URL count, lastmod dates, priority values |
-| HTML `<title>` | Page source | Record exact value |
-| `<meta description>` | Page source | Record exact value |
-| `<link rel="canonical">` | Page source | Record exact value |
-| Open Graph tags | Page source | og:title, og:description, og:type, og:image |
-| HTTP response headers | Server response | Server, Cache-Control, Content-Type, X-Robots-Tag, Strict-Transport-Security |
-| Page word count | Rendered HTML | After JS execution — Playwright renders dynamic content |
-| Heading hierarchy | Rendered HTML | H1 count, H2 count, nesting order |
-| Structured data | Page source | JSON-LD blocks, Microdata, RDFa |
-| HTTPS status | URL scheme + redirect | Confirm HTTPS redirect from HTTP |
-
-In agent-only mode, perform this extraction manually by reading the page via fetch/curl equivalent and parsing visible content.
-
----
-
-## STEP 5 — PHASE 2: SUBAGENT DISPATCH LOGIC
-
-After Phase 1 completes, dispatch all 5 subagents in parallel. Each subagent receives the Phase 1 data as input.
-
-### geo-ai-visibility (25% of GEO Score)
-
-**Scope:** AI crawler access and citation format quality
-
-**Inputs:** robots.txt parsed rules, llms.txt contents (if any), page HTML structure
-
-**What it checks:**
-- Permission status for each of the 14 AI crawlers (GPTBot, ClaudeBot, PerplexityBot, Google-Extended, Bingbot, Applebot-Extended, Anthropic-AI, cohere-ai, Meta-ExternalFetcher, YouBot, DuckAssistBot, Scrapy, CCBot, ia_archiver)
-- Whether llms.txt exists and follows the correct format
-- Whether llms-full.txt exists
-- Citation format quality — does the page structure support clean citation by AI systems?
-- Presence of clean canonical URLs, no JavaScript-only content walls
-
-**Outputs:** Crawler permission matrix (14 rows), llms.txt status, citation format score (0–100)
-
----
-
-### geo-content (20% of GEO Score)
-
-**Scope:** E-E-A-T signals, answer block quality, content self-containment
-
-**Inputs:** Rendered page HTML, word count, heading structure, author markup
-
-**What it checks:**
-- Experience signals: case studies, original data, first-person examples
-- Expertise signals: author credentials, bylines, About page depth
-- Authoritativeness signals: external citations, backlink signals (if available), Wikipedia mentions
-- Trustworthiness signals: HTTPS, contact info, privacy policy, review count
-- Answer block quality: does the page contain self-contained answer paragraphs that AI systems can quote directly?
-- Self-containment check: word count vs. pronoun density — pages with high pronoun density relative to noun density score lower
-- Statistical density: presence of percentages, numbered lists, data references
-- Uniqueness signals: original research, proprietary data, non-generic claims
-
-**Outputs:** E-E-A-T score (0–100), content quality score (0–100), top 3 content gaps
-
----
-
-### geo-platform-analysis (10% of GEO Score)
-
-**Scope:** Readiness for specific AI search platforms
-
-**Inputs:** Page structure, schema markup, llms.txt status, content signals
-
-**What it checks:**
-- ChatGPT readiness: Browse with Bing access, clean URL structure, answer-block format
-- Perplexity readiness: Direct URL crawlability, citation format, structured answers
-- Google AI Overviews readiness: E-E-A-T signals, structured data, FAQ schema, featured snippet optimization
-- Gemini readiness: Google entity graph signals, Knowledge Panel readiness, schema breadth
-
-**Outputs:** Platform readiness table (4 platforms × 5 signals × score), composite platform score (0–100)
-
----
-
-### geo-schema (10% of GEO Score)
-
-**Scope:** Structured data coverage and validation
-
-**Inputs:** Page source JSON-LD blocks, Microdata, RDFa markup
-
-**What it checks:**
-- Which schema types are present (Organization, WebSite, Article, BreadcrumbList, FAQPage, HowTo, Product, LocalBusiness, etc.)
-- Validation errors (missing required properties, incorrect value types)
-- Missing recommended schema types for the page type
-- Schema richness score — number of types × completeness per type
-
-**Outputs:** Schema type coverage table, validation error list, missing types list, schema score (0–100)
-
----
-
-### geo-technical (15% of GEO Score)
-
-**Scope:** Technical SEO and server health
-
-**Inputs:** HTTP response headers, HTTPS status, page load signals, robots.txt
-
-**What it checks:**
-- HTTPS enforcement (redirect from HTTP, HSTS header)
-- Mobile friendliness signals (viewport meta tag, responsive CSS indicators)
-- robots.txt validity (valid syntax, no wildcard blocks on critical paths)
-- Sitemap accessibility (sitemap.xml found, linked from robots.txt)
-- Core Web Vitals signals (from Lighthouse or page source — LCP hints, CLS hints, FID/INP hints)
-- Server headers (Cache-Control, Content-Encoding, CDN indicators)
-- Page speed indicators (image optimization signals, script loading strategy)
-
-**Outputs:** Server headers table, Core Web Vitals signal table, technical score (0–100), critical fixes list
-
----
-
-## STEP 6 — GEO SCORE SYNTHESIS
-
-Apply the weighted formula after all subagent results are collected.
-
-### GEO Score Formula
-
-```
-GEO Score = (AI Citability × 0.25) + (Brand Authority × 0.20) + (Content Quality × 0.20) +
-            (Technical Foundations × 0.15) + (Structured Data × 0.10) + (Platform Optimization × 0.10)
-```
-
-### Component Score Normalization
-
-Each component subagent returns a raw score 0–100. If a component cannot be scored (data gap), use 50 as the neutral default and flag the component as `data-gap`.
-
-### Letter Grade Thresholds
-
-| Grade | Score Range | Meaning |
-|---|---|---|
-| A | 85–100 | Highly optimized for AI search — strong citability, clean access, rich schema |
-| B | 70–84 | Good AI visibility — some gaps, addressable in one sprint cycle |
-| C | 55–69 | Moderate visibility — missing key citability signals, schema gaps, crawler issues |
-| D | 40–54 | Poor AI visibility — likely blocked or thin content, major remediation needed |
-| F | Below 40 | Not AI-search-ready — crawlers blocked, no schema, low citability |
-
-### Score Interpretation for Client Communication
-
-- **A (85+):** "Your site is well-positioned for AI-driven search. We recommend ongoing monitoring and targeted improvements to maintain this position."
-- **B (70–84):** "Good foundation. Targeted improvements in [lowest component] can push you into the A tier within 30–60 days."
-- **C (55–69):** "Moderate visibility. You are not capturing significant AI-referred traffic yet. A full remediation roadmap is recommended."
-- **D (40–54):** "Your site has significant AI search blindspots. AI systems may not be citing or recommending your content at all."
-- **F (below 40):** "Critical issues detected. AI crawlers may be blocked outright. Immediate technical and content remediation required."
-
----
-
-## STEP 7 — CITABILITY ALGORITHM
-
-The citability score measures how well a page can be cleanly quoted and cited by AI systems.
-
-### 5-Metric Algorithm
-
-| Metric | Weight | What It Measures |
-|---|---|---|
-| Answer Block Quality | 30% | Do paragraphs contain complete, self-sufficient answers AI can quote verbatim? |
-| Self-Containment | 25% | Can a paragraph be understood without reading surrounding context? Low pronoun-to-noun ratio = better. |
-| Structural Readability | 20% | Does the page use headings, short paragraphs, and numbered lists to enable AI parsing? |
-| Statistical Density | 15% | Does the page contain specific numbers, percentages, and data references AI prefers to cite? |
-| Uniqueness Signals | 10% | Does the content contain proprietary claims, original research, or data not found elsewhere? |
-
-### Computing Each Metric
-
-- **Answer Block Quality:** Score each paragraph 0–10 on: clear subject, predicate, supporting evidence, no unresolved pronouns. Average across all paragraphs.
-- **Self-Containment:** Count pronoun references (it, they, this, that) as a fraction of total noun references. Lower ratio = higher score.
-- **Structural Readability:** Check presence of: H1 (required), H2s (≥2), numbered/bulleted lists (≥1), paragraph length ≤150 words, no wall-of-text sections.
-- **Statistical Density:** Count numerical data references (% signs, numbered lists, dates with years, dollar figures) per 1000 words. Optimal: 8–15 data points per 1000 words.
-- **Uniqueness Signals:** Check for: "first-party study," "based on our data," "internal research," proprietary methodology references, or unique terminology.
-
----
-
-## STEP 8 — REMEDIATION ROADMAP LOGIC
-
-Build the roadmap by ranking all findings by impact × urgency.
-
-### Priority Matrix
-
-| Priority | Impact | Urgency | Action |
-|---|---|---|---|
-| P0 — Critical | High | Immediate | Fix this week — crawlers blocked, HTTPS broken, major schema errors |
-| P1 — High | High | This sprint | Add llms.txt, fix citability blockers, add missing schema types |
-| P2 — Medium | Medium | Week 2–3 | E-E-A-T improvements, statistical density, answer block rewrites |
-| P3 — Low | Low | Week 4+ | Brand authority campaigns, platform optimization, monitoring setup |
-
-### 4-Week Sprint Structure
-
-- **Week 1:** P0 and P1 items — all technical and crawler-access blockers
-- **Week 2:** P2 items — content quality improvements, schema additions
-- **Week 3:** P2 continuation — E-E-A-T depth, self-containment rewrites
-- **Week 4:** P3 items — brand platform seeding, monitoring, reporting loop setup
-
----
-
-## STEP 9 — OUTPUT ORDER
-
-Produce artifacts in this strict order:
-
-1. GEO Audit Brief (`templates/geo-audit-brief.md`)
-2. Citability Analysis (`templates/citability-analysis.md`)
-3. Crawler Access Report (`templates/crawler-access-report.md`)
-4. Brand Visibility Report (`templates/brand-visibility-report.md`)
-5. GEO Score Summary (`templates/geo-score-summary.md`)
-6. Content Analysis (`templates/content-analysis.md`)
-7. Schema Validation (`templates/schema-validation.md`)
-8. Technical Foundations (`templates/technical-foundations.md`)
-9. llms.txt Plan (`templates/llmstxt-plan.md`)
-10. Remediation Roadmap (`templates/remediation-roadmap.md`)
-11. Client Proposal (`templates/client-proposal.md`) — only if requested
-
----
-
-## STEP 10 — QUALITY BAR
-
-Good output looks like this:
-
-- All scores derived from actual page data — no invented numbers
-- GEO Score uses the exact formula defined in Step 6 — no rounding before final aggregate
-- Citability analysis applies all 5 metrics, not just the top 2
-- Crawler access report covers all 14 AI crawlers by exact user-agent string
-- Remediation roadmap is sequenced by impact × urgency, not alphabetically
-- Client proposal grounds ROI projections in AI search traffic trend data
-- Every output file can be handed to a client or developer and acted on immediately
-- No filler paragraphs — every sentence either presents data, explains a finding, or specifies an action

package/assets/worker-kits/growthub-geo-seo-v1/templates/brand-visibility-report.md

@@ -1,101 +0,0 @@
-# Brand Visibility Report
-
-> Template: `templates/brand-visibility-report.md`
-> Save output to: `output/<client-slug>/<project-slug>/BrandVisibilityReport_v<N>_<YYYYMMDD>.md`
-
----
-
-## Brand Audited
-
-| Field | Value |
-|---|---|
-| Brand Name | <!-- client_name --> |
-| Primary Domain | <!-- https://domain.com --> |
-| Brand Slug | <!-- brand-slug --> |
-| Scan Date | <!-- YYYY-MM-DD --> |
-| Execution Mode | <!-- local-fork (brand_scanner.py) / agent-only --> |
-| Script Used | <!-- scripts/brand_scanner.py / manual --> |
-
----
-
-## Platform Scan Results
-
-| Platform | Brand Present | Mention Count | Sentiment | Primary URL Found | Notes |
-|---|---|---|---|---|---|
-| YouTube | <!-- yes / no / partial --> | <!-- N --> | <!-- positive / neutral / negative / mixed --> | <!-- https://... --> | <!-- channel / mentions in videos --> |
-| Reddit | <!-- yes / no / partial --> | <!-- N --> | <!-- positive / neutral / negative / mixed --> | <!-- subreddit or post URL --> | <!-- notable threads --> |
-| Wikipedia | <!-- yes / no / partial --> | <!-- N --> | <!-- neutral --> | <!-- https://... --> | <!-- article or mention in related article --> |
-| LinkedIn | <!-- yes / no / partial --> | <!-- N --> | <!-- positive / neutral / mixed --> | <!-- https://... --> | <!-- company page / employee mentions --> |
-| Twitter / X | <!-- yes / no / partial --> | <!-- N --> | <!-- positive / neutral / negative / mixed --> | <!-- https://x.com/... --> | <!-- handle / mentions --> |
-| GitHub | <!-- yes / no / partial --> | <!-- N --> | <!-- positive / neutral --> | <!-- https://github.com/... --> | <!-- repos / mentions --> |
-| Quora | <!-- yes / no / partial --> | <!-- N --> | <!-- positive / neutral / negative --> | <!-- https://... --> | <!-- answers mentioning brand --> |
-| Hacker News | <!-- yes / no / partial --> | <!-- N --> | <!-- positive / neutral / negative --> | <!-- https://news.ycombinator.com/... --> | <!-- posts / comments --> |
-
----
-
-## Platform Detail Notes
-
-### YouTube
-<!-- Describe what was found: official channel exists / brand mentioned in competitor content / no presence found -->
-
-### Reddit
-<!-- Describe key threads, subreddits where brand appears, sentiment breakdown -->
-
-### Wikipedia
-<!-- Note: own Wikipedia article / mentioned in industry article / no presence -->
-
-### LinkedIn
-<!-- Official company page status, follower count if visible, employee mention density -->
-
-### Twitter / X
-<!-- Official handle / mention volume / notable brand advocates or critics -->
-
-### GitHub
-<!-- Open source repos / developer tools / mentions in issues or discussions -->
-
-### Quora
-<!-- Questions answered by brand representatives / brand mentioned in answers -->
-
-### Hacker News
-<!-- Any "Show HN" posts / brand mentioned in discussions / competitor comparisons -->
-
----
-
-## Brand Authority Score
-
-| Component | Score (0–100) | Weight | Weighted Score | Notes |
-|---|---|---|---|---|
-| Platform presence breadth | <!-- N --> | 30% | <!-- N × 0.30 --> | <!-- # platforms with meaningful presence / 8 --> |
-| Sentiment quality | <!-- N --> | 25% | <!-- N × 0.25 --> | <!-- ratio of positive to negative mentions --> |
-| Mention volume | <!-- N --> | 20% | <!-- N × 0.20 --> | <!-- total mentions normalized vs. category benchmark --> |
-| Wikipedia or authority link | <!-- N --> | 15% | <!-- N × 0.15 --> | <!-- Wikipedia presence = 100, none = 0 --> |
-| Developer / technical community | <!-- N --> | 10% | <!-- N × 0.10 --> | <!-- GitHub + HN combined signal --> |
-| **TOTAL Brand Authority Score** | | **100%** | <!-- sum --> | |
-
----
-
-## Gap Analysis
-
-**Platforms with no meaningful presence:**
-- <!-- Platform: why it matters and what to build there -->
-- <!-- Platform: why it matters and what to build there -->
-
-**Sentiment risks:**
-- <!-- Any platform with negative-dominant sentiment: what is being said -->
-
-**Brand authority gaps vs. competitors:**
-| Gap | Competitor Advantage | Recommended Response |
-|---|---|---|
-| <!-- gap --> | <!-- competitor URL or example --> | <!-- recommended action --> |
-| <!-- gap --> | <!-- competitor URL or example --> | <!-- recommended action --> |
-
----
-
-## Recommended Actions
-
-| Priority | Platform | Action | Timeline | Expected Impact |
-|---|---|---|---|---|
-| P1 | <!-- platform --> | <!-- specific action --> | <!-- Week 1-2 --> | <!-- High / Medium --> |
-| P1 | <!-- platform --> | <!-- specific action --> | <!-- Week 1-2 --> | <!-- High / Medium --> |
-| P2 | <!-- platform --> | <!-- specific action --> | <!-- Week 2-4 --> | <!-- Medium --> |
-| P3 | <!-- platform --> | <!-- specific action --> | <!-- Week 4+ --> | <!-- Low / Long-term --> |

package/assets/worker-kits/growthub-geo-seo-v1/templates/citability-analysis.md

@@ -1,131 +0,0 @@
-# Citability Analysis
-
-> Template: `templates/citability-analysis.md`
-> Save output to: `output/<client-slug>/<project-slug>/CitabilityAnalysis_v<N>_<YYYYMMDD>.md`
-
----
-
-## URL Audited
-
-| Field | Value |
-|---|---|
-| Target URL | <!-- https://... --> |
-| Client | <!-- client_name --> |
-| Analysis Date | <!-- YYYY-MM-DD --> |
-| Execution Mode | <!-- local-fork / agent-only --> |
-| Script Used | <!-- scripts/citability_scorer.py / manual --> |
-
----
-
-## Citability Score
-
-| Metric | Value |
-|---|---|
-| Raw Score | <!-- 0–100 --> |
-| Letter Grade | <!-- A / B / C / D / F --> |
-| Percentile Estimate | <!-- e.g., "Top 28% of audited pages" --> |
-| Score Date | <!-- YYYY-MM-DD --> |
-| Previous Score | <!-- N/A or previous score + date --> |
-
----
-
-## Component Breakdown
-
-| Metric | Raw Score (0–100) | Weight | Weighted Score | Status |
-|---|---|---|---|---|
-| Answer Block Quality | <!-- score --> | 30% | <!-- score × 0.30 --> | <!-- strong / adequate / weak --> |
-| Self-Containment | <!-- score --> | 25% | <!-- score × 0.25 --> | <!-- strong / adequate / weak --> |
-| Structural Readability | <!-- score --> | 20% | <!-- score × 0.20 --> | <!-- strong / adequate / weak --> |
-| Statistical Density | <!-- score --> | 15% | <!-- score × 0.15 --> | <!-- strong / adequate / weak --> |
-| Uniqueness Signals | <!-- score --> | 10% | <!-- score × 0.10 --> | <!-- strong / adequate / weak --> |
-| **TOTAL** | | **100%** | <!-- sum --> | |
-
----
-
-## Answer Block Quality Detail
-
-**Score: <!-- N/A → fill -->**
-
-Paragraphs evaluated: <!-- count -->
-
-| Paragraph # | Subject Clear | Evidence Present | No Unresolved Pronouns | Score |
-|---|---|---|---|---|
-| 1 | <!-- yes/no --> | <!-- yes/no --> | <!-- yes/no --> | <!-- 0–10 --> |
-| 2 | <!-- yes/no --> | <!-- yes/no --> | <!-- yes/no --> | <!-- 0–10 --> |
-| 3 | <!-- yes/no --> | <!-- yes/no --> | <!-- yes/no --> | <!-- 0–10 --> |
-
-**Best answerable paragraph:**
-> <!-- quote the strongest paragraph verbatim -->
-
-**Weakest paragraph (needs rewrite):**
-> <!-- quote the weakest paragraph verbatim -->
-
----
-
-## Self-Containment Check
-
-| Signal | Value | Status |
-|---|---|---|
-| Total word count | <!-- N --> | <!-- good (>800) / low (<300) --> |
-| Pronoun count | <!-- N --> | |
-| Noun count | <!-- N --> | |
-| Pronoun-to-noun ratio | <!-- N:N --> | <!-- good (<0.3) / moderate / poor (>0.6) --> |
-| Optimal word count range | 800–2,500 words | |
-| Contextual dependency score | <!-- 0–100 --> | <!-- lower is better --> |
-
----
-
-## Structural Readability Check
-
-| Signal | Status | Notes |
-|---|---|---|
-| H1 present | <!-- yes / no --> | <!-- H1 text if present --> |
-| H2 count | <!-- N --> | <!-- ≥2 recommended --> |
-| H3 count | <!-- N --> | |
-| Numbered lists | <!-- yes / no --> | <!-- count --> |
-| Bulleted lists | <!-- yes / no --> | <!-- count --> |
-| Average paragraph length | <!-- N words --> | <!-- ≤150 recommended --> |
-| Wall-of-text sections | <!-- yes / no --> | <!-- describe if yes --> |
-
----
-
-## Statistical Density Check
-
-| Signal | Value | Status |
-|---|---|---|
-| Percentage references | <!-- N --> | |
-| Numbered data points | <!-- N --> | |
-| Dollar/currency figures | <!-- N --> | |
-| Year/date references | <!-- N --> | |
-| Data points per 1,000 words | <!-- N.N --> | <!-- optimal: 8–15 --> |
-| Statistical density score | <!-- 0–100 --> | |
-
----
-
-## Uniqueness Signals Check
-
-| Signal | Present | Example |
-|---|---|---|
-| First-party study or research | <!-- yes / no --> | <!-- excerpt --> |
-| Proprietary data references | <!-- yes / no --> | <!-- excerpt --> |
-| Original methodology | <!-- yes / no --> | <!-- excerpt --> |
-| Unique terminology | <!-- yes / no --> | <!-- excerpt --> |
-| Non-generic claims | <!-- yes / no --> | <!-- excerpt --> |
-
----
-
-## Key Findings
-
-1. <!-- Most impactful finding — what is driving the score up or down most -->
-2. <!-- Second finding -->
-3. <!-- Third finding -->
-
----
-
-## Top 3 Improvements
-
-| Priority | Action | Expected Score Gain | Effort |
-|---|---|---|---|
-| 1 | <!-- Specific action, e.g., "Rewrite intro paragraph to remove 4 unresolved pronouns" --> | +<!-- N --> points | <!-- Low / Medium / High --> |
-| 2 | <!-- Specific action --> | +<!-- N --> points | <!-- Low / Medium / High --> |
-| 3 | <!-- Specific action --> | +<!-- N --> points | <!-- Low / Medium / High --> |