Website Growth Intelligence Report

Overview

This skill produces a practical, client-ready growth intelligence report for a website, product page, creator profile, or business landing page. It uses FluxA paid APIs through the x402 proxy to collect rendered-page evidence, technical signals, broken-link risk, SEO/accessibility/security scores, web/news context, and optional visual assets.

The best default use case is a Website Growth Intelligence Report because it is broadly useful, low-cost, and combines the most reliable APIs in the starter kit:

• StorkTools for rendered scraping, website grading, headers, DNS/WHOIS, tech stack, screenshots, broken links, YouTube/Reddit research, and PDF output.
• Web Search API for current news/context around the brand, market, product, or competitor topic.
• GPT Image 2 optionally, only when the user wants a branded hero/thumbnail/cover image because it costs more than the other calls.
• PredictScope optionally for market-facing companies, tokens, public assets, prediction-market context, or trend intelligence.
• Alpha Beds only for travel/hospitality workflows; do not book anything without explicit user confirmation.

The final output should be a clean report with severity-ranked findings, evidence, recommendations, and a short execution roadmap.

When to Use

Use this skill when the user asks for any of the following:

• Audit a website, landing page, product page, funnel, or directory listing.
• Find SEO/GEO/AEO improvement opportunities for a brand or page.
• Check technical credibility: broken links, headers, tech stack, DNS/WHOIS, security and accessibility basics.
• Produce a public-facing report or PDF deliverable for a client, bounty, or proof URL.
• Research a company/brand and summarize current news/context plus website signals.
• Create a visual cover/thumbnail for an audit report, only if a paid image generation call is justified.

Do not use this skill for:

• Fully automated hotel booking or payment finalization. Alpha Beds can search and verify availability, but booking requires explicit user approval.
• Real-money or irreversible transactions without an explicit user request and a signed mandate.
• Medical, legal, or financial advice reports. You can collect signals, but label conclusions as informational.

FluxA API Inventory

All APIs are accessed through FluxA proxy URLs and require x402 payment for paid endpoints. Every paid request must include X-Agent-ID; use fluxa-wallet status to get the current agent ID.

GPT Image 2

• Base: https://proxy-monetize.fluxapay.xyz/api/gpt-image-2/c5ebab8a
• Endpoint: POST /v1/images/generations
• Price: 0.20 USDC
• Body fields:

- model string, required, example: gpt-image-2 - prompt string, required

• Best use: optional report cover, social thumbnail, infographic hero.

StorkTools

• Base: https://proxy-monetize.fluxapay.xyz/api/storktools/c5ebab8a
• Key endpoints:

- GET /grade?url=<url> — full website audit. Price: 0.08 USDC. - GET /check-links?url=<url>&timeout=10000 — broken link scan. Price: 0.05 USDC. - GET /scrape?url=<url> — rendered text scrape. Price: 0.03 USDC. - GET /screenshot?url=<url>&width=1440&height=1200&fullpage=true&format=png — screenshot. Price: 0.03 USDC. - GET /techstack?url=<url> — framework/CMS/analytics/CDN detection. Price: 0.03 USDC. - GET /headers?url=<url> — response headers. Price: 0.01 USDC. - GET /dns?domain=<domain>&type=ALL — DNS records. Price: 0.01 USDC. - GET /whois?domain=<domain> — registrar and domain age. Price: 0.01 USDC. - POST /markdown with { "url": "...", "includelinks": "true", "include_images": "false" } — rendered page to clean markdown. Price: 0.01 USDC. - POST /pdf with { "url": "...", "format": "A4" } — generate PDF. Price: 0.05 USDC. - GET /qr?data=<url>&size=512 — QR code. Price: 0.01 USDC. - Optional research: reddit/search, reddit/read, youtube/search, youtube/info, youtube/transcript, youtube/comments.

Web Search API

• Base: https://proxy-monetize.fluxapay.xyz/api/web-search-api/c5ebab8a
• Endpoint: POST /search-news?query=<query>
• Price: 0.01 USDC
• Best use: current company, competitor, market, launch, or incident context.

PredictScope

• Base: https://proxy-monetize.fluxapay.xyz/api/predictscope/c5ebab8a
• Endpoints:

- POST /mcp/v1/weave — read-only price/trend/news intelligence. Price: 0.01 USDC. - POST /mcp/v1/trading — paper trading with isolated virtual capital. Price: 0.01 USDC.

• Best use: optional market intelligence when the audited brand relates to public markets, crypto, commodities, prediction markets, or a tradable narrative.
• Start with tools/list before invoking named MCP tools.

Alpha Beds

• Base: https://proxy-monetize.fluxapay.xyz/api/alpha-beds/c5ebab8a
• Useful endpoints:

- POST /api/fluxa/v1/intent — natural-language hotel search. Price: 0.01 USDC. - POST /api/fluxa/v1/rooms — room availability. Price: 0.01 USDC. - POST /api/fluxa/v1/verify — real-time price/availability verification. Price: 0.01 USDC. - POST /api/fluxa/v1/create and /create/confirm — booking/payment flow; use only after explicit user approval.

• Best use in this skill: travel/hospitality competitive research or itinerary-adjacent reports, not autonomous booking.

Skill Ideas Considered

1. Website Growth Intelligence Report — best default. Low-cost, evidence-rich, and broadly useful for SEO/GEO, client audits, bounties, product launches, and lead generation.
2. Travel Booking Concierge — combines Alpha Beds, Web Search, StorkTools, and optional GPT Image, but booking creates higher risk and requires user confirmation.
3. Prediction Market Research Desk — uses PredictScope plus Web Search for market narratives; strong for trading simulations but narrower audience.
4. Social/Video Content Intelligence Brief — uses StorkTools YouTube/Reddit endpoints plus GPT Image for content planning; useful but less comprehensive.
5. Lead Qualification Bot — uses StorkTools techstack/WHOIS/headers and Web Search to score prospects; good for sales ops, but the website report is easier to present as a complete deliverable.

Chosen implementation: Website Growth Intelligence Report.

Output Standard

Return a professional report, not raw API dumps. Recommended structure:

# Website Growth Intelligence Report: <brand/domain>

## Executive Summary
- Overall verdict: <1 paragraph>
- Top 3 opportunities:
  1. ...
  2. ...
  3. ...
- Top 3 risks:
  1. ...
  2. ...
  3. ...

## Evidence Collected
- URL audited: <url>
- APIs used: <list>
- Timestamp: <UTC timestamp>
- Estimated API cost: <USDC>

## Scorecard
- Performance: <score or qualitative>
- SEO/GEO readiness: <score>
- Accessibility: <score>
- Security/trust: <score>
- Technical clarity: <score>
- Content clarity: <score>

## Findings by Severity
### Critical
- Finding: ...
- Evidence: ...
- Impact: ...
- Fix: ...

### High
...

### Medium
...

### Low
...

## GEO / AI Search Recommendations
- Entity clarity:
- Structured answers:
- Source credibility:
- FAQ/schema opportunities:
- Comparison/query capture:

## 7-Day Action Plan
- Day 1-2:
- Day 3-5:
- Day 6-7:

## Appendix: Raw Signals
- Headers summary:
- Tech stack summary:
- Broken links summary:
- News/context summary:

Cost-Aware Default Call Plan

Use the cheapest evidence-first plan unless the user asks for visuals or PDF export.

Minimal report, about 0.16 USDC

• StorkTools /grade: 0.08
• StorkTools /markdown: 0.01
• StorkTools /headers: 0.01
• StorkTools /techstack: 0.03
• StorkTools /scrape: 0.03

Standard report, about 0.28 USDC

• Everything in Minimal: 0.16
• StorkTools /check-links: 0.05
• StorkTools /screenshot: 0.03
• StorkTools /dns: 0.01
• StorkTools /whois: 0.01
• Web Search /search-news: 0.01
• StorkTools /qr: 0.01

Premium report, about 0.53+ USDC

• Everything in Standard: 0.28
• GPT Image 2 cover: 0.20
• StorkTools /pdf: 0.05
• Optional YouTube/Reddit/PredictScope calls as needed: 0.01-0.05 each

Payment and x402 Flow

1. Verify FluxA Wallet

fluxa-wallet status

If the CLI is missing:

npm install -g @fluxa-pay/fluxa-wallet

If no agent is configured:

fluxa-wallet init --name "Website Growth Auditor" --client "FluxA Website Growth Intelligence Report skill"

Extract the agent ID from fluxa-wallet status. Do not print JWTs, tokens, private keys, or secrets.

2. Create or Reuse a Mandate

Create a mandate with enough budget for the planned report. Atomic units use USDC 6 decimals: 100000 = 0.10 USDC, 500000 = 0.50 USDC, 1000000 = 1.00 USDC.

fluxa-wallet mandate-create \
  --desc "Spend up to 0.50 USDC for a website growth intelligence report" \
  --amount 500000 \
  --seconds 28800 \
  --category research

Open or share the returned authorizationUrl so the user can sign. Then verify:

fluxa-wallet mandate-status --id <mandate_id>

Continue only when mandate.status is signed.

3. Make a Paid API Call

First call the API without X-Payment to receive the HTTP 402 payload. Include tracking headers.

AGENT_ID="<agent_id>"
SKILL_NAME="website-growth-intelligence-report"
TASK_ID="audit-$(date +%Y%m%d%H%M%S)"
API_URL="https://proxy-monetize.fluxapay.xyz/api/storktools/c5ebab8a/grade?url=https%3A%2F%2Fexample.com"

curl -sS -i "$API_URL" \
  -H "X-Agent-ID: $AGENT_ID" \
  -H "X-Skill-Name: $SKILL_NAME" \
  -H "X-Task-Id: $TASK_ID" \
  -o /tmp/fluxa_402_response.txt

Extract the JSON body and filter the accepts array to the network matching the mandate currency. For USDC mandates, keep network == "base".

BODY=$(sed -n '/^\r$/,$p' /tmp/fluxa_402_response.txt | tail -n +2)
PAYLOAD=$(printf '%s' "$BODY" | jq '{x402Version: .x402Version, accepts: [.accepts[] | select(.network == "base")] }')

PAYMENT=$(fluxa-wallet x402-v3 --mandate <mandate_id> --payload "$PAYLOAD")
XPAYMENT=$(printf '%s' "$PAYMENT" | jq -r '.data.xPaymentB64')

curl -sS "$API_URL" \
  -H "X-Agent-ID: $AGENT_ID" \
  -H "X-Skill-Name: $SKILL_NAME" \
  -H "X-Task-Id: $TASK_ID" \
  -H "X-Payment: $XPAYMENT"

If using FLUXAMONETIZECREDITS, filter to network == "fluxa-monetize-credits" instead.

End-to-End Workflow

Step 1 — Normalize Inputs

Required input:

• url: the page or website to audit.

Optional inputs:

• brand_name
• primary_goal: examples: SEO, GEO, conversion, credibility, launch readiness, lead generation.
• competitors: 1-5 competitor URLs or names.
• audience: target buyers/users.
• budget_mode: minimal, standard, or premium.
• output_format: chat report, Markdown file, PDF, or public proof page.

Normalize the domain:

python - <<'PY'
from urllib.parse import urlparse
url = '<url>'
if not url.startswith(('http://','https://')):
    url = 'https://' + url
p = urlparse(url)
print({'url': url, 'domain': p.netloc.replace('www.', '')})
PY

Step 2 — Collect Website Signals

Use StorkTools calls according to the selected budget:

# Grade
GET https://proxy-monetize.fluxapay.xyz/api/storktools/c5ebab8a/grade?url=<encoded_url>

# Rendered markdown
POST https://proxy-monetize.fluxapay.xyz/api/storktools/c5ebab8a/markdown
{"url":"<url>","include_links":"true","include_images":"false"}

# Headers
GET https://proxy-monetize.fluxapay.xyz/api/storktools/c5ebab8a/headers?url=<encoded_url>

# Tech stack
GET https://proxy-monetize.fluxapay.xyz/api/storktools/c5ebab8a/techstack?url=<encoded_url>

# Broken links
GET https://proxy-monetize.fluxapay.xyz/api/storktools/c5ebab8a/check-links?url=<encoded_url>&timeout=10000

# Screenshot
GET https://proxy-monetize.fluxapay.xyz/api/storktools/c5ebab8a/screenshot?url=<encoded_url>&width=1440&height=1200&full_page=true&format=png

# DNS and WHOIS
GET https://proxy-monetize.fluxapay.xyz/api/storktools/c5ebab8a/dns?domain=<domain>&type=ALL
GET https://proxy-monetize.fluxapay.xyz/api/storktools/c5ebab8a/whois?domain=<domain>

Step 3 — Collect Market and News Context

Use Web Search API with a concise query:

POST https://proxy-monetize.fluxapay.xyz/api/web-search-api/c5ebab8a/search-news?query=<brand_or_topic>

Good queries:

• <brand name> company product launch funding
• <domain> reviews complaints security
• <category> AI agent payments competitors
• <brand> alternatives pricing

Step 4 — Optional Social/Video Context

For content-heavy brands, use StorkTools YouTube/Reddit endpoints:

GET /youtube/search?query=<brand_or_topic>&max_results=5
GET /youtube/info?video_id=<id>
GET /youtube/transcript?video_id=<id>&lang=en
GET /reddit/search?query=<brand_or_topic>&sort=relevance&time_filter=year&limit=10
GET /reddit/read?url=<reddit_thread_url>&limit=50

Summarize recurring objections, praise, confusion points, and exact user language.

Step 5 — Optional Market Intelligence with PredictScope

Use this only if the brand/topic is connected to markets or prediction narratives.

Start by listing tools:

POST https://proxy-monetize.fluxapay.xyz/api/predictscope/c5ebab8a/mcp/v1/weave
{"jsonrpc":"2.0","method":"tools/list","id":"1"}

Then call the appropriate tool using JSON-RPC. Keep this read-only unless the user explicitly asks for paper trading.

Step 6 — Optional Visual Asset

Use GPT Image 2 only when a visual deliverable is requested. Create a specific prompt based on audit findings:

POST https://proxy-monetize.fluxapay.xyz/api/gpt-image-2/c5ebab8a/v1/images/generations
{
  "model": "gpt-image-2",
  "prompt": "Professional dark-mode report cover for a website growth intelligence audit of <brand>. Visual themes: search visibility, AI agents, trust signals, technical diagnostics, modern SaaS dashboard. No logos unless provided."
}

Step 7 — Synthesize the Report

Do not paste raw JSON. Convert evidence into prioritized recommendations.

Severity rubric:

• Critical: blocks indexing, conversion, trust, access, payment, or legal compliance.
• High: materially hurts SEO/GEO visibility, user trust, performance, or conversion.
• Medium: meaningful improvement opportunity with moderate impact.
• Low: polish, hygiene, or future optimization.

For each finding include:

• finding
• evidence
• impact
• recommended_fix
• owner (marketing, engineering, design, content, ops)
• estimated_effort (S/M/L)
• severity

Step 8 — Deliver Files

If the user needs a file:

• Write Markdown to a local .md file.
• Optionally use StorkTools /pdf for a PDF export.
• If proof is needed, publish to the user's preferred public host and verify the URL.

Example Report Prompt

Use this synthesis prompt after data collection:

Create a Website Growth Intelligence Report for <brand/domain>.
Primary goal: <goal>.
Audience: <audience>.
Evidence available:
- StorkTools grade summary: <summary>
- Rendered markdown/page copy: <summary>
- Headers/security signals: <summary>
- Tech stack: <summary>
- Broken links: <summary>
- DNS/WHOIS: <summary>
- Web/news context: <summary>

Requirements:
- Professional, client-ready Markdown.
- Start with an executive summary.
- Include severity-ranked findings.
- Include GEO/AI-search recommendations.
- Include a 7-day action plan.
- Every recommendation must cite evidence from collected signals.
- Do not hallucinate scores if not returned by the APIs; label qualitative estimates clearly.

Travel/Hospitality Variant

If the site is a hotel, event, travel agency, or destination brand, optionally use Alpha Beds for real-time hotel intelligence:

POST https://proxy-monetize.fluxapay.xyz/api/alpha-beds/c5ebab8a/api/fluxa/v1/intent
{"query":"Find hotels near <destination> for <dates>, <guests>, under <budget>"}

Use results to benchmark location, price, availability, room language, and offer clarity. Do not call /create or /create/confirm unless the user explicitly asks to book and approves the exact hotel, room, date, guest details, cancellation policy, and price.

Common Pitfalls

1. Calling paid endpoints without X-Agent-ID. Discovery works without it, but paid endpoint calls require tracking headers.
2. Passing the full multi-network accepts array to x402-v3. Filter to the mandate currency: base for USDC or fluxa-monetize-credits for credits.
3. Forgetting that GPT Image 2 costs more. Do not use it in minimal reports unless the user asks for a visual asset.
4. Treating Alpha Beds as safe to auto-book. Search and verify are okay under a mandate; booking requires explicit user approval.
5. Dumping raw API output. Users need ranked findings and fixes, not JSON blobs.
6. Overclaiming live facts. Use Web Search/news results as evidence and include timestamps.
7. Leaking credentials. Never print JWTs, tokens, private keys, or full secret-bearing config files.
8. Ignoring failed or partial calls. If one API fails, continue with available evidence and label missing sections.

Verification Checklist

• AGENTS.md and FluxA wallet docs were read before making changes.
• API discovery was run for GPT Image 2, StorkTools, Web Search API, Alpha Beds, and PredictScope.
• The selected workflow has a clear use case and combines at least two FluxA APIs.
• The call plan includes estimated costs and avoids unnecessary high-cost calls.
• x402 payment instructions include mandate creation, 402 payload handling, accepts filtering, and retry with X-Payment.
• Every paid API request includes X-Agent-ID, X-Skill-Name, and X-Task-Id.
• The final report format includes evidence, severity, impact, fixes, and action plan.
• No secrets, JWTs, API tokens, or private keys are included in the output.