The-AI-Directory-Company/agents-and-skills / skills/on-page-sgeo
on-page-sgeo
VerifiedOptimize individual pages for both search engine ranking and AI citation — covering title tags, meta descriptions, heading hierarchy, URL structure, internal linking, image optimization, structured data, direct-answer formatting, and knowledge block structure.
| File | Kind | Size |
|---|---|---|
| SKILL.md | skill md | 25.2 KB |
| references/geo-formatting.md | markdown | 10.4 KB |
| references/heading-and-structure.md | markdown | 5.8 KB |
| references/image-and-media.md | markdown | 5.7 KB |
| references/internal-linking.md | markdown | 6.3 KB |
| references/meta-optimization.md | markdown | 5.6 KB |
| scripts/analyze-headings.pyexecutable | script | 5.9 KB |
| scripts/audit-page.pyexecutable | script | 14.6 KB |
| scripts/check-direct-answer.pyexecutable | script | 9.1 KB |
| scripts/check-freshness.pyexecutable | script | 12.1 KB |
| scripts/check-images.pyexecutable | script | 7.3 KB |
| scripts/check-internal-links.pyexecutable | script | 7.6 KB |
| scripts/extract-meta-tags.pyexecutable | script | 7.1 KB |
| scripts/extract-structured-data.pyexecutable | script | 8.1 KB |
scripts/check-direct-answer.py
View on GitHub#!/usr/bin/env python3
"""Check the direct-answer quality of a page's opening content.
Extracts the first 200 words after the H1, analyzes for direct-answer
patterns vs meandering intros, checks for anaphoric references, and
scores on the 0-8 GEO per-section rubric from references/geo-formatting.md.
Scoring criteria (0-2 each, total 0-8):
- Direct answer present
- Self-contained
- Specific data
- Source attribution
Usage:
python check-direct-answer.py --url https://example.com/page
python check-direct-answer.py --url https://example.com/page --tools tools.json
Output: JSON with first 200 words, score breakdown, issues, and patterns found.
"""
import argparse
import json
import re
import sys
from html.parser import HTMLParser
from typing import Optional
# ---------------------------------------------------------------------------
# Content extractor
# ---------------------------------------------------------------------------
class ContentExtractor(HTMLParser):
"""Extract text content from the main/article area after the first H1."""
SKIP_TAGS = {"nav", "header", "footer", "aside", "script", "style", "noscript"}
def __init__(self):
super().__init__()
self._found_h1 = False
self._after_h1 = False
self._skip_depth = 0
self._text_parts: list[str] = []
self._in_main = False
self._main_depth = 0
def handle_starttag(self, tag: str, attrs):
if tag in ("main", "article"):
self._in_main = True
self._main_depth += 1
if tag in self.SKIP_TAGS:
self._skip_depth += 1
if tag == "h1" and not self._found_h1:
self._found_h1 = True
elif self._found_h1 and not self._after_h1 and tag != "h1":
self._after_h1 = True
def handle_endtag(self, tag: str):
if tag in ("main", "article"):
self._main_depth -= 1
if self._main_depth <= 0:
self._in_main = False
if tag in self.SKIP_TAGS and self._skip_depth > 0:
self._skip_depth -= 1
def handle_data(self, data: str):
if self._after_h1 and self._skip_depth == 0:
self._text_parts.append(data)
def get_text(self) -> str:
raw = " ".join(self._text_parts)
return re.sub(r"\s+", " ", raw).strip()
# ---------------------------------------------------------------------------
# Analysis patterns
# ---------------------------------------------------------------------------
MEANDERING_PATTERNS = [
(r"\bIn today'?s\b", "In today's..."),
(r"\bAs we all know\b", "As we all know..."),
(r"\bWhen it comes to\b", "When it comes to..."),
(r"\bIt'?s no secret that\b", "It's no secret that..."),
(r"\bIn this (comprehensive|complete|ultimate) guide\b", "In this comprehensive guide..."),
(r"\bBefore we dive in\b", "Before we dive in..."),
(r"\bLet'?s take a step back\b", "Let's take a step back..."),
(r"\bThe world of .+ has changed\b", "The world of X has changed..."),
(r"\bWelcome to\b", "Welcome to..."),
(r"\bWe'?re (proud|excited|thrilled) to\b", "We're proud/excited to..."),
]
ANAPHORIC_PATTERNS = [
(r"\bAs mentioned above\b", "As mentioned above"),
(r"\bAs we discussed\b", "As we discussed"),
(r"\bThis approach\b", "This approach"),
(r"\bThe above\b", "The above"),
(r"\bSee (the )?previous section\b", "See previous section"),
(r"\bUsing the same (method|approach|technique)\b", "Using the same method"),
(r"\bBuilding on this\b", "Building on this"),
(r"^It\b", "Starts with 'It' (potentially anaphoric)"),
]
# Patterns indicating specific data
DATA_PATTERNS = [
r"\d+%", # Percentages
r"\$[\d,.]+", # Dollar amounts
r"\d{4}", # Years
r"\d+\.\d+", # Decimal numbers
r"\d+,\d{3}", # Thousands
r"\d+ (seconds?|minutes?|hours?|days?|weeks?|months?|years?)", # Durations
r"\d+x\b", # Multipliers
]
# Patterns indicating source attribution
SOURCE_PATTERNS = [
r"according to",
r"(a |the )?\d{4} .+ (study|survey|report|analysis|research)",
r"(Ahrefs|Semrush|Moz|HubSpot|Gartner|Forrester|McKinsey|Deloitte|Google|Bing)",
r"(study|survey|report) (of|by|from) \d+",
r"\b(published|reported) (in|by)\b",
]
def get_first_n_words(text: str, n: int = 200) -> str:
words = text.split()
return " ".join(words[:n])
def find_patterns(text: str, patterns: list[tuple[str, str]]) -> list[str]:
found = []
for regex, label in patterns:
if re.search(regex, text, re.IGNORECASE):
found.append(label)
return found
def count_matches(text: str, patterns: list[str]) -> int:
count = 0
for p in patterns:
count += len(re.findall(p, text, re.IGNORECASE))
return count
# ---------------------------------------------------------------------------
# Scoring (0-8 rubric)
# ---------------------------------------------------------------------------
def score_section(text: str) -> dict:
"""Score the first 200 words on the 0-8 GEO per-section rubric."""
first_200 = get_first_n_words(text, 200)
words = first_200.split()
meandering = find_patterns(first_200, MEANDERING_PATTERNS)
anaphoric = find_patterns(first_200, ANAPHORIC_PATTERNS)
data_count = count_matches(first_200, DATA_PATTERNS)
source_count = count_matches(first_200, SOURCE_PATTERNS)
# --- Direct answer present (0-2) ---
if meandering:
direct_answer = 0
elif len(words) < 20:
direct_answer = 0 # Too short to evaluate
else:
# Heuristic: if no meandering patterns and text starts with a
# substantive statement (not a question), likely direct.
first_sentence = first_200.split(".")[0] if "." in first_200 else first_200
if len(first_sentence.split()) >= 8:
direct_answer = 2
else:
direct_answer = 1
# --- Self-contained (0-2) ---
if len(anaphoric) >= 2:
self_contained = 0
elif len(anaphoric) == 1:
self_contained = 1
else:
self_contained = 2
# --- Specific data (0-2) ---
if data_count >= 3:
specific_data = 2
elif data_count >= 1:
specific_data = 1
else:
specific_data = 0
# --- Source attribution (0-2) ---
if source_count >= 2:
source_attribution = 2
elif source_count >= 1:
source_attribution = 1
else:
source_attribution = 0
total = direct_answer + self_contained + specific_data + source_attribution
issues = []
if direct_answer < 2:
issues.append("Opening does not start with a direct answer")
if self_contained < 2:
issues.append("Contains anaphoric references that reduce standalone readability")
if specific_data < 2:
issues.append("Lacks specific data points (numbers, percentages, dates)")
if source_attribution < 2:
issues.append("No named source citations in opening")
return {
"first_200_words": first_200,
"word_count": len(words),
"score": total,
"breakdown": {
"direct_answer": direct_answer,
"self_contained": self_contained,
"specific_data": specific_data,
"source_attribution": source_attribution,
},
"interpretation": (
"citation_ready" if total >= 6
else "partially_citable" if total >= 3
else "not_citable"
),
"issues": issues,
"meandering_patterns": meandering,
"anaphoric_references": anaphoric,
}
# ---------------------------------------------------------------------------
# Fetch
# ---------------------------------------------------------------------------
def fetch_html(url: str) -> str:
from urllib.request import Request, urlopen
req = Request(url, headers={"User-Agent": "Mozilla/5.0 (compatible; on-page-sgeo-audit/1.0)"})
with urlopen(req, timeout=15) as resp:
return resp.read().decode("utf-8", errors="replace")
# ---------------------------------------------------------------------------
# Main
# ---------------------------------------------------------------------------
def main():
ap = argparse.ArgumentParser(
description="Check direct-answer quality and GEO readiness of a page's opening content"
)
ap.add_argument("--url", required=True, help="Page URL to analyze")
ap.add_argument("--tools", default=None, help="Path to tools.json inventory file")
args = ap.parse_args()
try:
html = fetch_html(args.url)
except Exception as e:
print(json.dumps({"error": f"Failed to fetch URL: {e}", "url": args.url}), file=sys.stderr)
sys.exit(1)
extractor = ContentExtractor()
extractor.feed(html)
text = extractor.get_text()
if not text:
print(json.dumps({
"url": args.url,
"error": "Could not extract text content after H1",
"score": 0,
}))
sys.exit(0)
result = score_section(text)
result["url"] = args.url
print(json.dumps(result, indent=2))
if __name__ == "__main__":
main()
Related skills
| Skill | Stars | Updated | Works with | Status | Install |
|---|---|---|---|---|---|
| slack-gif-creatoranthropics/skillsKnowledge and utilities for creating animated GIFs optimized for Slack. Provides constraints, validation tools, and animation concepts. Use when users request animated GIFs for Slack like "make me a GIF of X doing Y for Slack." | Stars: 198k | Updated | Claude CodeCodex | Verified | Install |
| ponytailDietrichGebert/ponytailLazy senior dev mode for any coding task: the smallest change that fully solves it, and a reply a busy human understands in one read. | Stars: 196k | Updated | Claude CodeCodex | Verified | Install |
| ponytail-debtDietrichGebert/ponytailHarvest every shortcut: comment into one debt ledger, so deferrals get tracked instead of forgotten. One-shot report. | Stars: 196k | Updated | Claude CodeCodex | Verified | Install |
| ponytail-debtDietrichGebert/ponytailList every `shortcut:` comment (and older `ponytail:` ones) as a debt ledger. One-shot report, changes nothing. Use for "ponytail debt", "what did ponytail defer", "list the shortcuts", /ponytail-debt. | Stars: 196k | Updated | Claude CodeCodex | Verified | Install |