Bhanunamikaze/Agentic-SEO-Skill
seo
VerifiedDeterministic LLM-first SEO audits for websites, blog posts, and GitHub repositories. Use this when the user asks to "perform SEO analysis", "run SEO audit", "analyze SEO", "check technical SEO", "review schema", "Core Web Vitals", "E-E-A-T", "hreflang", "GEO", "AEO", or GitHub repository SEO optimization. For full/page/repo audits, run bundled scripts for evidence and return prioritized, confidence-labeled fixes.
scripts/github_readme_lint.py
View on GitHub#!/usr/bin/env python3
"""
GitHub README SEO Lint
Scores README quality for GitHub discoverability and conversion readiness.
Usage:
python github_readme_lint.py README.md --json
"""
import argparse
import base64
import json
import os
import re
import sys
from datetime import datetime, timezone
from github_api import GitHubAPIError, fetch_json, get_token, resolve_repo
DEFAULT_INTENTS = [
"seo",
"audit",
"technical seo",
"schema",
"core web vitals",
]
def utc_now_iso() -> str:
return datetime.now(timezone.utc).replace(microsecond=0).isoformat()
def read_text(path: str) -> str:
with open(path, "r", encoding="utf-8") as f:
return f.read()
def looks_like_placeholder(text: str) -> bool:
value = (text or "").strip().lower()
if value in {"404", "404: not found", "not found"}:
return True
if value.startswith("404:"):
return True
if value.startswith("<html") and "not found" in value:
return True
return False
def fetch_readme_from_repo(repo: str, token: str, provider: str) -> str:
response = fetch_json(f"/repos/{repo}/readme", token=token, provider=provider, timeout=30)
payload = response.get("data", {})
content = payload.get("content") or ""
if not content:
return ""
raw = base64.b64decode(content.encode("utf-8"), validate=False)
return raw.decode("utf-8", errors="replace")
def strip_code_fences(text: str) -> str:
text = re.sub(r"(?ms)^```[\s\S]*?^```[ \t]*$", "", text)
text = re.sub(r"(?ms)^~~~[\s\S]*?^~~~[ \t]*$", "", text)
return text
def extract_headings(markdown: str) -> list:
headings = []
lines = markdown.splitlines()
for i, line in enumerate(lines, start=1):
m = re.match(r"^(#{1,6})\s+(.+?)\s*$", line.strip())
if m:
headings.append({"line": i, "level": len(m.group(1)), "text": m.group(2).strip()})
continue
# Setext headings:
# Title
# =====
if i < len(lines):
next_line = lines[i].strip()
if line.strip() and re.match(r"^(=+|-+)\s*$", next_line):
level = 1 if next_line.startswith("=") else 2
headings.append({"line": i, "level": level, "text": line.strip()})
return headings
def count_code_blocks(markdown: str) -> int:
fenced_backticks = len(re.findall(r"(?ms)^```[\s\S]*?^```[ \t]*$", markdown))
fenced_tildes = len(re.findall(r"(?ms)^~~~[\s\S]*?^~~~[ \t]*$", markdown))
fenced_total = fenced_backticks + fenced_tildes
# Count indented code blocks (>=2 consecutive indented non-empty lines)
indented_total = 0
run = 0
for line in markdown.splitlines():
if re.match(r"^( |\t)\S", line):
run += 1
else:
if run >= 2:
indented_total += 1
run = 0
if run >= 2:
indented_total += 1
return fenced_total + indented_total
def extract_images(markdown: str) -> list:
images = []
for m in re.finditer(r"!\[(.*?)\]\((.*?)\)", markdown):
images.append({"alt": (m.group(1) or "").strip(), "url": (m.group(2) or "").strip()})
return images
def plain_word_count(markdown: str) -> int:
text = strip_code_fences(markdown)
text = re.sub(r"!\[.*?\]\(.*?\)", " ", text)
text = re.sub(r"\[([^\]]+)\]\([^)]+\)", r"\1", text)
text = re.sub(r"[#>*`_~\-]", " ", text)
words = re.findall(r"\b[\w']+\b", text.lower())
return len(words)
def add_finding(findings: list, category: str, severity: str, finding: str, evidence: str, fix: str):
findings.append(
{
"category": category,
"severity": severity,
"confidence": "Confirmed",
"finding": finding,
"evidence": evidence,
"fix": fix,
}
)
def normalize_heading_text(headings: list) -> list:
return [h["text"].strip().lower() for h in headings]
def detect_heading_jumps(headings: list) -> list:
jumps = []
prev = None
for h in headings:
if prev is not None and h["level"] > prev + 1:
jumps.append({"line": h["line"], "from": prev, "to": h["level"], "text": h["text"]})
prev = h["level"]
return jumps
def score_report(markdown: str, intents: list) -> dict:
headings = extract_headings(markdown)
heading_text = normalize_heading_text(headings)
images = extract_images(markdown)
words = plain_word_count(markdown)
findings = []
# Opening clarity (20)
opening_score = 20
opening_chunk = strip_code_fences(markdown).strip().splitlines()[:20]
opening_text = " ".join(opening_chunk).lower()
matched_intents = [term for term in intents if term.lower() in opening_text]
if not matched_intents:
opening_score -= 12
add_finding(
findings,
"Opening Clarity",
"Warning",
"Opening section lacks target intent terms.",
"None of the configured intent terms appear in the opening section.",
"Include primary use-case language in first 2-3 paragraphs.",
)
if words < 250:
opening_score -= 4
add_finding(
findings,
"Opening Clarity",
"Info",
"README is short for a discoverability-oriented project page.",
f"Word count is {words}.",
"Expand with concise sections for install, proof, and contribution paths.",
)
opening_score = max(0, opening_score)
# Information architecture (20)
ia_score = 20
h1_count = sum(1 for h in headings if h["level"] == 1)
heading_jumps = detect_heading_jumps(headings)
if h1_count != 1:
ia_score -= 12
severity = "Critical" if h1_count == 0 else "Warning"
add_finding(
findings,
"Information Architecture",
severity,
"README should contain exactly one H1 heading.",
f"Detected H1 count: {h1_count}.",
"Keep a single H1 title and move other top-level sections to H2.",
)
if heading_jumps:
ia_score -= 5
add_finding(
findings,
"Information Architecture",
"Warning",
"Heading hierarchy has level jumps.",
f"Detected {len(heading_jumps)} jump(s) where heading level skips intermediary levels.",
"Normalize heading flow (H1 -> H2 -> H3) without skipping levels.",
)
ia_score = max(0, ia_score)
# Install + quickstart (20)
install_score = 20
install_markers = ("install", "quick start", "quickstart", "getting started", "setup")
has_install_section = any(any(marker in t for marker in install_markers) for t in heading_text)
if not has_install_section:
install_score -= 14
add_finding(
findings,
"Install + Quickstart",
"Critical",
"Installation/quickstart section is missing.",
"No install or getting-started heading detected.",
"Add a dedicated install section with copy-paste commands and prerequisites.",
)
code_block_count = count_code_blocks(markdown)
if code_block_count == 0:
install_score -= 4
add_finding(
findings,
"Install + Quickstart",
"Warning",
"No code examples detected.",
"README contains zero detectable code blocks (fenced or indented).",
"Add runnable command examples for setup and core usage.",
)
install_score = max(0, install_score)
# Proof + credibility (15)
proof_score = 15
proof_markers = ("example", "output", "report", "screenshot", "demo", "result")
has_examples_section = any(any(marker in t for marker in proof_markers) for t in heading_text)
if not has_examples_section:
proof_score -= 8
add_finding(
findings,
"Proof + Credibility",
"Warning",
"README lacks explicit proof/results section.",
"No heading found for examples, reports, screenshots, or outputs.",
"Add evidence sections with sample outputs or screenshots.",
)
if "license" not in " ".join(heading_text) and "license" not in markdown.lower():
proof_score -= 4
add_finding(
findings,
"Proof + Credibility",
"Warning",
"License reference is missing or unclear in README.",
"No explicit license mention detected.",
"Add a short license section linking to LICENSE file.",
)
proof_score = max(0, proof_score)
# CTA + community (15)
cta_score = 15
cta_markers = ("contribut", "issue", "pull request", "support", "discussion", "star")
cta_hits = sum(1 for marker in cta_markers if marker in markdown.lower())
has_contributing_section = "contribut" in markdown.lower()
if cta_hits < 2:
cta_score -= 9
add_finding(
findings,
"CTA + Community",
"Warning",
"README has weak contribution/support call-to-action coverage.",
f"Detected {cta_hits} CTA marker(s) from contribution/support keyword set.",
"Add clear paths for contributing, opening issues, and support requests.",
)
cta_score = max(0, cta_score)
# Readability + accessibility (10)
read_score = 10
missing_alt = [img for img in images if not img["alt"]]
if missing_alt:
read_score -= 6
add_finding(
findings,
"Readability + Accessibility",
"Warning",
"Some README images are missing alt text.",
f"{len(missing_alt)} image(s) detected with empty alt text.",
"Add descriptive alt text for each non-decorative image.",
)
if len(headings) < 4:
read_score -= 2
add_finding(
findings,
"Readability + Accessibility",
"Info",
"README sectioning is shallow.",
f"Detected only {len(headings)} heading(s).",
"Use additional headings to improve scannability and structure.",
)
read_score = max(0, read_score)
category_scores = {
"opening_clarity": opening_score,
"information_architecture": ia_score,
"install_quickstart": install_score,
"proof_credibility": proof_score,
"cta_community": cta_score,
"readability_accessibility": read_score,
}
total_score = sum(category_scores.values())
if total_score >= 90:
rating = "Excellent"
elif total_score >= 70:
rating = "Good"
elif total_score >= 50:
rating = "Needs Improvement"
elif total_score >= 30:
rating = "Poor"
else:
rating = "Critical"
if not findings:
add_finding(
findings,
"Overall",
"Pass",
"README meets baseline GitHub SEO and conversion quality checks.",
"No major structural, install, or accessibility deficiencies detected.",
"Maintain quality with periodic linting and updates.",
)
return {
"summary": {"score": total_score, "rating": rating},
"metrics": {
"word_count": words,
"heading_count": len(headings),
"h1_count": h1_count,
"code_block_count": code_block_count,
"image_count": len(images),
"images_missing_alt": len(missing_alt),
"matched_intents_in_opening": matched_intents,
"has_install_section": has_install_section,
"has_examples_section": has_examples_section,
"has_contributing_section": has_contributing_section,
},
"category_scores": category_scores,
"findings": findings,
}
def print_text(report: dict):
summary = report.get("summary", {})
print("\nREADME SEO Lint")
print("=" * 60)
print(f"Score: {summary.get('score', 'NA')}/100 ({summary.get('rating', 'Unknown')})")
metrics = report.get("metrics", {})
print(
f"Words: {metrics.get('word_count', 0)} | Headings: {metrics.get('heading_count', 0)} | "
f"Images missing alt: {metrics.get('images_missing_alt', 0)}"
)
print("\nTop findings:")
for f in report.get("findings", [])[:10]:
print(f"- [{f['severity']}] {f['finding']}")
def main():
parser = argparse.ArgumentParser(description="README SEO and conversion linting for GitHub repositories.")
parser.add_argument("readme_path", nargs="?", default="README.md", help="Path to README file (default: README.md)")
parser.add_argument("--repo", help="Repository slug or URL (owner/repo) for remote README fallback.")
parser.add_argument("--provider", choices=["auto", "api", "gh"], default="auto", help="GitHub data provider mode for README fallback.")
parser.add_argument("--token", help="GitHub token override. Prefer env vars GITHUB_TOKEN or GH_TOKEN.")
parser.add_argument("--intent", action="append", help="Target intent phrase (repeatable).")
parser.add_argument("--json", action="store_true", help="Output JSON.")
parser.add_argument("--output", help="Write JSON report to file path.")
args = parser.parse_args()
intents = args.intent if args.intent else DEFAULT_INTENTS
limitations = []
content_source = "local_file"
markdown = ""
if os.path.exists(args.readme_path):
markdown = read_text(args.readme_path)
else:
limitations.append(f"Local README file not found: {args.readme_path}")
needs_remote = (not markdown.strip()) or looks_like_placeholder(markdown)
if needs_remote and args.repo:
try:
token = get_token(args.token)
repo = resolve_repo(args.repo)
remote_md = fetch_readme_from_repo(repo=repo, token=token, provider=args.provider)
if remote_md.strip():
markdown = remote_md
content_source = "github_api"
limitations.append("Used remote README fallback due missing/placeholder local README.")
except GitHubAPIError as exc:
limitations.append(f"Remote README fallback failed: {exc}")
if looks_like_placeholder(markdown):
print(
"Error: README content appears to be placeholder/404. Provide valid README path or repo access for remote fallback.",
file=sys.stderr,
)
sys.exit(2)
if not markdown.strip():
print("Error: README content unavailable from local file and remote fallback.", file=sys.stderr)
sys.exit(2)
scored = score_report(markdown=markdown, intents=intents)
report = {
"timestamp_utc": utc_now_iso(),
"readme_path": args.readme_path,
"content_source": content_source,
"limitations": limitations,
"intents": intents,
**scored,
}
if args.output:
with open(args.output, "w", encoding="utf-8") as f:
json.dump(report, f, indent=2)
if args.json:
print(json.dumps(report, indent=2))
else:
print_text(report)
if __name__ == "__main__":
main()
Related skills
| Skill | Stars | Updated | Works with | Status | Install |
|---|---|---|---|---|---|
| rea-analysis-optimizationmorluto/reaOptimize REA providers and shared analysis workflows for agent context, latency, CPU, and memory while preserving correctness and Evidence. Use for analysis bottlenecks and cache or query design; CI/test-suite optimization belongs to rea-ci-tests-audit. | Stars: 471k | Updated | Claude CodeCodex | Verified | Install |
| rea-ci-tests-auditmorluto/reaAudit REA CI and test suites for correctness, fixture drift, false passes, and measured performance; use for suite or workflow investigations and efficiency changes, rather than ordinary feature testing or a single test failure. | Stars: 471k | Updated | Claude CodeCodex | Verified | Install |
| rea-tool-designmorluto/reaAudit, investigate, design, or change REA investigation tools, CLI/MCP contracts, provider capabilities, and Evidence semantics. | Stars: 471k | Updated | Claude CodeCodex | Verified | Install |
| ponytail-auditDietrichGebert/ponytailQuality audit of the whole repo: bugs, security, real load, missing tests, speed, and what to delete. Most important first. | Stars: 194k | Updated | Claude CodeCodex | Verified | Install |