Content Audit Skill — Setup Guide

What this does

Audits your content library using a Job / Evidence / Gap framework. Connects to your analytics, pulls real data, auto-flags everything, and gives you a prioritized CSV with actions.

Quick start

1. Install dependencies

You need Python 3 and requests:

pip3 install requests

If using Google Search Console or GA4, also install:

pip3 install google-auth google-api-python-client  # for GSC
pip3 install google-analytics-data google-auth      # for GA4

2. Connect your analytics

Pick your platform and set up the connection.

PostHog

You need three things:

• API key: Go to PostHog → Settings → Personal API Keys → Create. Give it read-only scopes for Query, Insight, Event definition, and Person.
• Project ID: It's in your PostHog URL: posthog.com/project/<THIS_NUMBER>
• Host: https://us.posthog.com (US) or https://eu.posthog.com (EU)

Set them in your terminal:

export POSTHOG_API_KEY="phx_your_key_here"
export POSTHOG_PROJECT_ID="12345"
export POSTHOG_HOST="https://us.posthog.com"

Google Search Console

You need a Google Cloud service account:

1. Go to Google Cloud Console → create a project (or use existing)
2. Enable the "Search Console API" in APIs & Services → Library
3. Go to APIs & Services → Credentials → Create Credentials → Service Account
4. Click the service account → Keys tab → Add Key → JSON (downloads a .json file)
5. Go to Google Search Console → your property → Settings → Users and permissions → Add user → paste the service account email (the @...iam.gserviceaccount.com one) → set to Full

Set in your terminal:

export GSC_CREDENTIALS_PATH="/path/to/your-credentials.json"

Google Analytics 4

Same service account setup as GSC, but:

1. Enable the "Analytics Data API" (not the regular Analytics API)
2. In GA4 → Admin → Property → Property Access Management → add the service account email as Viewer
3. Your property ID is in GA4 → Admin → Property Settings (just the number)

Set in your terminal:

export GA4_CREDENTIALS_PATH="/path/to/your-credentials.json"

3. Pull your data

First, see what's available:

python3 scripts/pull_posthog_data.py --discover
python3 scripts/pull_gsc_data.py --discover --site-url "https://yoursite.com"
python3 scripts/pull_ga4_data.py --discover --property-id "123456789"

Then pull the CSV:

# PostHog
python3 scripts/pull_posthog_data.py --output audit-data.csv --url-pattern "/blog%" --conversion-events "sign_up,download"

# GSC
python3 scripts/pull_gsc_data.py --output audit-data.csv --site-url "https://yoursite.com" --url-filter "/blog/"

# GA4
python3 scripts/pull_ga4_data.py --output audit-data.csv --property-id "123456789" --url-filter "/blog/"

Replace the conversion events and URL filters with your own.

4. Run the audit

Option A: Warp

If you use Warp, install the .skill file and prompt: "Run a content audit on my blog content."

Option B: Claude / ChatGPT / any AI tool

1. Create a project (or custom GPT, or system prompt)
2. Add SKILL.md and references/framework.md as context
3. Attach your audit-data.csv
4. Prompt: "Run a content audit on this CSV"

The instructions will guide the AI to ask you about your product and ICP before starting.

5. What you get

• A flagged CSV with every piece of content rated RED / YELLOW / GREEN
• Grouped summary by content type
• Deep analysis on your top 20-30 pieces
• Gap analysis: what to create next, what to scale, what to kill