What does this skill do, and when should you use it?
Blog Audit is a full-site health assessment skill within the claude-blog suite. It scans project blog files and evaluates content quality, on-page SEO, structured data, link health, freshness, and AI citation readiness. It also detects topic cannibalization, orphan pages, dead-end pages, and stale content, then produces a prioritized action queue. It fits Claude Code users who need a batch review of locally available blog content.
Scans common blog directories for .md, .mdx, and .html files while excluding README, CHANGELOG, LICENSE, configuration files, SKILL.md, package.json, and node_modules. It uses the Task tool to run parallel analyses for content quality, on-page SEO, BlogPosting and FAQ schema, internal-link graphs, update dates, and AI citation readiness. It detects competing topics, orphan pages, dead-end pages, and posts older than 90 days, then writes a report containing a dashboard, per-post scores, issue counts, recommended link sources, refresh priorities, and estimated effort to blog-audit-report.md.
- A solo blogger wants to review every article against a consistent quality and SEO checklist.1
- A marketing team needs to identify keyword cannibalization, orphan pages, and outdated statistics across a content library.
- A content operator is planning a refresh cycle and needs prioritized posts and estimated refresh effort.
- A blog team wants one audit covering traditional SEO and AI citation readiness.
How do you install this skill?
- Execution scans project files and writes blog-audit-report.md at the project root, but overwrite behavior, confirmation, and recovery are unspecified.
- The required Task-based parallel subagents, permission boundaries, retry behavior, and result-merging rules are not defined in this skill.
- Quality, readability, and AI-citation scores are heuristic indicators, not proof of search ranking, factual accuracy, or citation impact.
- Date calculation, link reachability, HTML/MDX parsing, and Chinese-language support have no explicit limits or verification guidance.
- Repository context contains version and test-count inconsistencies; verify the pinned revision's actual contents before relying on them.
- Shell / CLI
- Local filesystem
Task tool
This skill is distributed as part of the public claude-blog plugin. In Claude Code 1.0.33+, run /plugin marketplace add AgriciDaniel/claude-blog, then /plugin install claude-blog@agricidaniel-claude-blog. The README does not document a separate installation procedure for blog-audit.
tmp="$(mktemp -d)"
git clone --depth 1 https://github.com/AgriciDaniel/claude-blog.git "$tmp"
mkdir -p ~/.claude/skills
cp -R "$tmp/skills/blog-audit" ~/.claude/skills/
rm -rf "$tmp"Generated from the source repository and skill path; it copies only this skill's folder. If the author's install steps above differ, follow those first. To scope it to one project, replace ~/.claude/skills with that project's .claude/skills.
How do you use this skill?
Once installed, send your agent any of these to trigger it:
- audit all posts,
- check all blogs.
In a project containing blog files, say “audit blog,” “blog audit,” “site audit,” “blog health,” “audit all posts,” or “check all blogs.” You can also run /blog audit [directory]. The skill scans the project and saves the complete report as blog-audit-report.md in the project root.
What are this skill's strengths and limitations?
- Covers content quality, SEO, structured data, links, freshness, and AI citation readiness in one pass.
- Finds keyword cannibalization, orphan pages, and pages with no outbound links.
- Produces per-post scores and a prioritized action queue.
- Supports common blog directories and falls back to a project-wide frontmatter search.
- Requires local project files and the Task tool; it is not documented as a standalone audit of an inaccessible live website.
- The source provides no blog-audit-specific test results or accuracy measurements.
- Its quality, readability, and AI-readiness scores are rule-based assessments, not evidence of search rankings or actual AI-platform citations.
How does this skill compare with similar options?
Side by side with related skills; every score comes from the same FSRS standard.
| Skill | FS score | Stars | Last updated | License |
|---|---|---|---|---|
| Blog Audit: Full-Site Health Review this page | 41 · Not recommended | ★ 2.4k | 14d ago | MIT |
| GEO Content Quality & E-E-A-T Assessment | 47 · Use with care | ★ 11k | 22d ago | MIT |
| Full-Site SEO Audit | 53 · Use with care | ★ 19k | 23d ago | MIT |
| Blog Analyzer | 51 · Use with care | ★ 2.4k | 14d ago | MIT |
| GEO Crawlers Access Analysis Skill | 51 · Use with care | ★ 11k | 22d ago | MIT |
How did FollowSkills review this skill?
The skill scopes discovery to blog files and explicitly filters configuration, dependency, and non-blog files, showing basic data-boundary awareness. It does not explain isolation of untrusted content, require confirmation before writing the report, provide rollback, or disclose parallel-agent data flow, so points are deducted.
The audit stages, abnormal empty-project branch, and report structure are detailed, and repository context includes CI and test evidence. However, the skill depends on an unspecified Task tool and lacks defined handling or feedback for parse errors, permission failures, invalid dates, and very large projects, so the static score is conservative.
Trigger phrases, the directory argument, intended use cases, and primary output are clear. Non-fit cases, semantic false-trigger boundaries, Chinese-language behavior, and mainland-China reachability are not specified; some quality metrics are language-biased, so points are deducted.
The skill has a readable staged structure, filtering rules, report templates, and MIT metadata, while repository context shows CI and maintenance signals. The selected skill itself lacks dependency installation notes, version/change information, maintenance ownership, and troubleshooting guidance; shared materials also contain v1.9.0/v1.9.1 and test-count inconsistencies, so points are deducted.
The intended result is an actionable site-wide report covering per-post scores, orphan pages, cannibalization, stale content, and prioritized actions. Static review cannot verify scan accuracy, scoring correctness, agent coordination, or direct usability of the generated report, so the score is limited by the static ceiling and uncertainty.
The skill specifies auditable discovery steps and a fixed report shape, and the repository context provides CI and committed test material. It contains no skill-specific real output, complete key-path execution evidence, or independent corroboration, so only limited credit is warranted.
Open a dimension to read why it scored that way
Evidence confidence：Low — Mostly static review, author material or a limited demo; useful for discovery, not high-risk decisions.
See the full review method →