How to Automate Your Law Firm Blog Audit with Claude Code

Most law firm blogs are an untapped liability.

Not because the content is wrong. Because nobody has checked it in years. Firms have posts published in 2021 referencing statutes that have since changed, articles missing attorney attribution, pages with thin content that Google suppresses and that AI answer engines like ChatGPT, Perplexity, and Google AI Overviews refuse to cite, and internal linking gaps bleeding authority away from their highest-value practice area pages.

Reviewing 50 or 100 blog posts by hand is not realistic for a practicing attorney. I know this because I’ve been doing law firm SEO for 15 years, and a thorough content audit used to take my team the better part of a week.

Now I run one in a few hours with a tool called Claude Code.

This guide walks you through the entire process, step by step. You don’t need a programming background. You need a computer, about 30 minutes to set things up, and blog content you’re willing to actually improve.

What Is Claude Code?

Claude Code is a tool built by Anthropic, the AI safety company behind the Claude AI family of models. According to Anthropic’s official documentation, Claude Code is “an agentic coding tool that reads your codebase, edits files, runs commands, and integrates with your development tools.”

It runs in your computer’s terminal and understands plain English instructions. You don’t write code with it. You talk to it, and it does the work.

For a law firm blog audit, you’ll use Claude Code to read your blog posts, follow a detailed audit checklist, flag issues, and produce a structured report. It can analyze dozens of posts in the time it would take you to carefully read three or four.

Claude Code requires a Claude subscription: Pro, Max, Teams, or Enterprise. Pricing is at claude.com/pricing. The Pro plan at $20/month is sufficient for the work in this guide.

What Makes a Legal Blog Audit Different

Not all content audits are created equal. Legal websites fall under what Google’s quality guidelines call YMYL content: Your Money or Your Life. These are pages where low-quality or inaccurate information could cause real harm.

Google’s E-E-A-T standards (Experience, Expertise, Authoritativeness, Trustworthiness) apply more strictly to legal content than to, say, a food blog or a travel site. A generic content audit tool won’t check for attorney author attribution, bar-compliant language, jurisdiction specificity, or citation of legal experience. This walkthrough is built specifically around what matters for law firm content.

Here’s what this audit covers:

E-E-A-T Signals

• Named attorney author with credentials on every post
• Publication and “last reviewed” dates
• References to specific legal experience and case types
• Jurisdiction-specific content where applicable
• Attorney advertising compliance flags (no uncertified “specialist” claims)

Legal Currency

• References to statutes, rules, or deadlines that may have changed
• Content from 2020-2022 still presented as current law
• Dollar thresholds or time limits typically adjusted annually
• COVID-era language presented as ongoing

Content Quality

• Word count below 800 words (thin for legal YMYL content)
• Posts with no H2 heading structure
• Missing FAQ coverage for practice area topics
• Duplicate or near-duplicate content across posts

Internal Linking

• Posts that don’t link to your main practice area pages
• Related blog posts that aren’t connected to each other
• Orphaned posts receiving no internal links

Technical Metadata

• Missing or duplicate meta descriptions
• Title tags over 60 characters
• Missing image alt text where present in content files

Step 1: Install Claude Code

Open your terminal. On a Mac, press Command + Space, type Terminal, and press Enter. On Windows, use PowerShell.

Run this installation command:

curl -fsSL https://claude.ai/install.sh | bash

According to Anthropic’s documentation, native installations automatically update in the background. You won’t need to reinstall it to stay current.

Alternatively, on a Mac with Homebrew installed:

brew install --cask claude-code

Note: Homebrew installations don’t auto-update. Run brew upgrade claude-code periodically.

After installation, start Claude Code:

claude

On first use you’ll be prompted to log in with your Claude account. Follow the on-screen instructions, then return here.

Step 2: Connect Claude to Your Content

In early 2024, the only option for this kind of audit was exporting blog posts to text files and dropping them in a local folder. That workflow still works - but there is now a faster and more powerful approach.

In November 2024, Anthropic introduced the Model Context Protocol (MCP), an open standard for connecting Claude Code directly to external data sources. Claude Code’s official documentation describes it this way: “With MCP, Claude Code can read your design docs in Google Drive, update tickets in Jira, pull data from Slack, or use your own custom tooling.” WordPress fits the same pattern.

The MCP approach (recommended for firms with developer support): Ask your developer to configure a read-only MCP server connection to your WordPress site. Once that integration is in place, Claude Code can query your CMS directly - reading posts, metadata, publish dates, and author data in real time without any file exports. The audit runs against your live content, which means no sync problems and no conversion steps.

If your firm uses a headless CMS (Webflow, Contentful, Sanity), the same principle applies - your developer configures the MCP connection and Claude Code reads the API.

The manual approach (fallback for firms without MCP setup): This still works well for most small to mid-sized firm blogs.

Ask your developer or marketing team to export your blog posts as plain text or Markdown files into a single folder. Most WordPress sites support this via the WP All Export plugin or WP-CLI.

For a quick manual export with fewer than 20 posts: Go to WordPress admin, navigate to Tools > Export > Posts, download the XML file, and ask your developer to convert it to individual text files. Or copy-paste each post body directly into its own .txt file.

File naming matters: Name files clearly so the audit output is readable.

• denver-car-accident-attorney-guide.txt
• colorado-dui-penalties.txt
• what-to-do-after-a-slip-and-fall.txt

Put all files in one folder. For this tutorial, the folder is blog-audit on your Desktop.

Step 3: Start Claude Code in Read-Only Mode

Here’s an important detail: when you’re doing an audit, you want Claude Code to read and report only. You don’t want it editing any files.

Claude Code has a built-in Plan Mode that restricts it to read-only analysis. Start it this way:

cd ~/Desktop/blog-audit
claude --permission-mode plan

The --permission-mode plan flag tells Claude Code to analyze and plan only. It will not modify, delete, or create files unless you explicitly switch modes. This is the right setting for an audit.

You’ll see the Claude Code interface. You’re ready to begin.

Step 4: Orient Claude Code to Your Firm

Before running audit checks, give Claude Code the context it needs to flag issues specific to legal content. Paste this at the prompt (editing for your situation):

I'm auditing blog content for a law firm. We practice [your practice areas, e.g., personal injury, criminal defense, family law] in [your city and state]. We follow [your state bar, e.g., Colorado Bar Association] advertising guidelines and the standards set out in ABA Formal Opinion 512 (2024) regarding competent use of generative AI. Our primary goals are: (1) improve E-E-A-T signals, (2) identify outdated or legally inaccurate content, (3) flag any case law citations or statute references that require independent human verification, and (4) identify content that needs updating or expansion. Please familiarize yourself with all the text files in this directory before we begin. When you're ready, confirm how many posts you see and give me a one-sentence summary of the topics covered.

Claude Code will scan your files and confirm the scope. Once it responds, move to the audit steps.

Step 5: E-E-A-T Signal Audit

This is your highest-priority check. Content strategy for competitive legal markets depends on E-E-A-T being built into every post, not bolted on after the fact. Paste this prompt:

Review all blog posts in this directory for E-E-A-T signals. For each post, check:
1. Is there a named attorney author with professional credentials listed?
2. Does the post include a publication date or "last reviewed" date?
3. Does the content reference specific legal experience, case types handled, or outcomes achieved (with appropriate disclaimers)?
4. Is jurisdiction-specific information included where relevant (state laws, local courts, filing deadlines)?
5. Does any language make uncertified "specialist" claims or guarantees that could violate bar advertising rules?
6. Does the post cite any specific court cases or statute sections? List every citation found - these will require independent human verification per ABA Formal Opinion 512 standards.

Return a table with four columns: Post Title or Filename | E-E-A-T Score (1-5) | Biggest E-E-A-T Gap | Priority Fix.
Sort by score, lowest first. After the table, list all case law and statute citations found across all posts in a separate section labeled "Citations Requiring Human Verification."

When Claude Code returns the table, ask a follow-up:

Which three posts have the lowest E-E-A-T scores and what are the exact sentences or sections I should rewrite first?

This gives you a specific to-do list, not just a score.

Step 6: Legal Currency and Accuracy Check

Legal information has an expiration date. A post from 2022 discussing Colorado’s comparative fault rules, federal student loan forgiveness, or state-specific COVID liability protections may be presenting outdated information as current fact.

Run this check:

Review all blog posts for content that may be legally outdated or time-sensitive. Look for:
- Specific year references (2019, 2020, 2021, 2022, 2023) suggesting content hasn't been reviewed recently
- Named statutes, regulations, or court rules that are frequently amended (e.g., Colorado Revised Statutes sections, Federal Rules of Civil Procedure)
- Dollar thresholds that are adjusted annually (e.g., workers' comp rates, filing fees, damage caps, mileage rates)
- Statutes of limitations or filing deadlines presented as definitive without noting they should be verified
- COVID-era language presented as current policy (e.g., court closures, remote hearings as default)
- Any claim presented as current law where the content is more than 18 months old

For each concern, list: Filename | Flagged Content (quote it) | Reason for Flag | Urgency (High/Medium/Low)

Claude Code is not a lawyer and cannot verify whether a statute has actually changed since your post was written. What it does is surface every post with date-sensitive content, giving you a targeted review list rather than reading everything from scratch.

The Case Law Citation Check

Run this as a separate, focused prompt immediately after the currency check:

Now audit specifically for legal citations across all posts. For each post:
1. List every named court case mentioned (e.g., "Plaintiff v. Defendant" or case shorthand)
2. List every statute or regulation cited by section number (e.g., "C.R.S. § 13-21-111")
3. List every regulatory body, agency rule, or administrative code referenced

For each citation found, return: Post Filename | Citation Type (Case / Statute / Regulation) | Exact Citation Text | Context (one sentence surrounding it)

Do not attempt to verify whether these citations are accurate. Just identify and list them all.

Why this matters: The 2023 Mata v. Avianca case - where attorneys submitted AI-generated fake case citations to a federal court - made the legal community acutely aware that AI tools can hallucinate case law. The ABA responded with Formal Opinion 512 (2024), which establishes that attorneys using generative AI tools have a specific duty to independently verify AI-generated legal research before relying on it.

This prompt doesn’t ask Claude Code to verify anything. It creates the list your attorney team needs to verify. That separation - AI surfaces, human confirms - is exactly what Opinion 512 requires.

Step 7: Thin Content Identification

Google’s quality rater guidelines define thin content as pages that provide little original value to users. For YMYL legal content, thin is particularly dangerous. Posts under 700-800 words on substantive legal topics rarely provide enough depth to satisfy a searcher’s genuine need - and Google suppresses them in rankings.

But in 2026 the penalty is double. Thin content doesn’t just lose rankings. It doesn’t get cited by the inference models powering Google AI Overviews, SearchGPT, or Perplexity. If your post on Colorado comparative fault law is 450 words with no FAQ section and no attorney attribution, no AI answer engine is going to surface it as a credible source when someone asks “who’s at fault in a multi-car pileup in Colorado?” You lose the ranking AND the citation. The gap compounds.

Run this:

For each blog post in this directory, estimate the following:
1. Approximate word count
2. Number of H2 headings (sections)
3. Whether there is a FAQ section (yes/no)
4. Whether there is a call to action (yes/no)

Return a table sorted by estimated word count, shortest first. Flag any post under 800 words and any post with fewer than two H2 headings.

Posts that come up thin usually fall into one of three categories: they were written as brief news items (can be updated or consolidated), they’re duplicate in substance to another post on the same topic (candidates for consolidation), or they’re legitimately narrow topics that don’t need more depth (acceptable).

One of the most common law firm SEO mistakes is keeping thin posts that compete with and dilute your stronger content on the same topic. Claude Code will surface those conflicts for you.

Step 8: Internal Linking Gap Analysis

Every blog post that doesn’t link to your main practice area pages is a missed opportunity to pass authority where it matters most. Most law firm blogs have significant gaps here, especially older content published before a deliberate linking strategy was in place.

Run this:

Review all blog posts for internal linking patterns. I want you to:
1. Identify which posts contain no outgoing links at all
2. Identify posts that discuss [insert your practice areas, e.g., personal injury, car accidents, criminal defense] but don't appear to link to a corresponding practice area page
3. Identify topics that appear in multiple posts that could be cross-linked (e.g., two posts about DUI that don't reference each other)
4. Suggest the top 10 specific internal linking additions that would most improve content connectivity

For each suggestion, note: Source Post | Suggested Anchor Text | Target Topic/Page

The output here is a direct action list. For each gap identified, one person adds one internal link inside your CMS. It’s the kind of work that’s tedious to identify manually but trivial to execute once you have the list.

Step 9: Meta Description Audit

If your exported files include frontmatter or metadata (common with Markdown exports or headless CMS setups), run this check:

For each blog post, check if a meta description is included in the file's frontmatter or header. For each post that has one:
1. Count the characters. Is it between 120 and 160 characters?
2. Does it include a geographic term (city, state) or practice area keyword?
3. Does it include a reason to click (question, benefit, or urgency)?

List posts with missing, too-short, or too-long meta descriptions. For the three worst, suggest a revised meta description.

If your content files don’t include meta descriptions (common with basic WordPress exports), note that and plan to run a separate meta description review inside your CMS using a tool like the built-in Yoast or RankMath interface.

Step 10: Generate the Consolidated Report

After running the individual checks, ask Claude Code to compile everything into one document:

Based on all the analysis you've run in this session, generate a comprehensive blog audit report with these sections:

1. Executive Summary (5 sentences maximum: what you found, severity, top recommendation)
2. E-E-A-T Issues: Top 5 posts to fix, what to fix, and in what order
3. Outdated Content Flags: Full list by urgency
4. Thin Content Priority List: Posts to expand or consolidate, sorted by opportunity
5. Internal Linking Action List: Top 10 additions with source, anchor text, and target
6. Quick Wins: Changes that can be completed inside the CMS in under 30 minutes each
7. 90-Day Update Roadmap: Suggested sequence for making all fixes

Then write this report to a file called blog-audit-report.txt in this directory.

That last line is key. Because you’re in plan mode, Claude Code will show you the report and confirm before creating the file. Review it, then type “yes” or “approve” to allow the file to be written.

When you open blog-audit-report.txt in any text editor, you’ll have a complete audit document you can share with your marketing team, SEO agency, or web developer.

What To Do With the Results

An audit without action is a document that collects dust. Here’s how I prioritize fixes for law firm clients after a blog audit:

This week - Quick Wins

Add attorney author attribution and publication dates to every post missing them. In most WordPress setups, this takes about two minutes per post. It’s the single highest-impact E-E-A-T fix available and requires no new writing.

Month 1 - High Priority Updates

Update the five most outdated posts identified in Step 6. Expand the three thinnest posts that cover high-value topics. Add internal links from your ten most-visited posts to your corresponding practice area pages.

Months 2-3 - Strategic Work

Consolidate posts on duplicate topics. Commission new content to fill the gaps the audit identified. Run a second audit to measure improvement against your baseline.

The AI-first approach to attorney marketing treats your existing content as a data problem. You can’t improve what you haven’t measured. The audit gives you the measurement.

A Note on What Claude Code Can and Can’t Do

Claude Code is a tool, not a lawyer, and this matters for legal content work.

It can identify patterns: missing attribution, date-sensitive language, thin word counts, absent FAQ sections, missing links. These are structural and textual signals it detects reliably.

It cannot verify whether a specific statute or regulation has been amended since your post was published. It cannot confirm that a named court case is real, accurately cited, or that the holding it describes is correct. It cannot provide a legal opinion on whether your content is bar-compliant in your jurisdiction.

The ABA addressed this directly. ABA Formal Opinion 512 (2024) established that attorneys using generative AI tools have a duty of competence that includes independently verifying AI-generated outputs before relying on them. This applies to using AI in client matters - but the underlying principle is sound for any legal content workflow: AI identifies, humans verify.

Use Claude Code’s output as a first-pass triage that surfaces what needs human review. You or your attorney team make the final calls on legal accuracy and bar compliance.

This is exactly why understanding AI’s evolving role in legal marketing matters right now. AI handles the mechanical audit work. You apply professional judgment to the results.

The Competitive Reality

Here’s what’s actually happening in 2026: potential clients are asking ChatGPT and Google Gemini complex legal questions and getting direct answers. They’re not always clicking through to websites anymore. The click-through rate on legal searches has dropped as AI Overviews absorb the answers.

That sounds like bad news for law firm SEO. It’s actually an opportunity - if your content is what those AI systems are drawing from.

When someone asks “What is the statute of limitations for a personal injury claim in Colorado?” the answer comes from somewhere. It comes from the content that AI inference models determined was authoritative, accurately sourced, and attorney-attributed. If your firm’s content is thin, outdated, or missing E-E-A-T signals, your firm doesn’t get the citation. If your competitor’s does, their name appears in the AI answer. That’s not just a lost click. That’s a lost brand impression at the exact moment someone decided they needed a lawyer.

GEO and traditional SEO are now inseparable for law firms. The same content fixes that improve Google rankings - author attribution, legal accuracy, depth, structured answers - are exactly what improves AI Overview citation rate, Perplexity sourcing, and ChatGPT references. A blog audit improves your standing across all of those channels at once.

Why law firms need specialized expertise for this work comes down to the checklist itself. The criteria for auditing legal content are different from auditing a retail site or SaaS blog. YMYL classification, bar compliance, ABA opinion standards, E-E-A-T requirements for attorney-authored content - none of these appear in generic content audit templates.

Going Deeper With Claude Code

This tutorial covers the blog audit use case. If you want to build more sophisticated AI workflows into your marketing operation, the Claude Code for Marketers course covers additional applications: automating competitive keyword research, generating structured content briefs at scale, and building custom audit templates for different practice area types. For firms ready to go beyond analysis and into deploying fully autonomous agents for intake, scheduling, or internal research, OpenClaw for law firm workflow automation picks up where this tutorial ends.

The command-line basics you learned here transfer directly to all of those workflows.

Ready to Run Your Audit?

You now have a complete playbook: install Claude Code, export your content, run eight specific audit checks, and compile a structured action report.

If you’d rather have a specialist handle the audit and execution rather than running it yourself, get in touch. My team has run these audits for law firms across every major practice area and can turn the results around faster with context your CMS data alone doesn’t provide.

Or if you’re specifically in the Denver market and want to talk through your firm’s content and SEO situation, I work directly with a small number of consulting clients - the details are on the Denver attorney SEO page.

Either way, don’t let your blog run another year without a structured review. The content you published two or three years ago is either compounding your authority or quietly undermining it. The audit tells you which.