Content Audit Pipeline

A content audit pipeline helps you evaluate existing content assets — blog posts, documentation, marketing pages — and identify what needs updating, consolidating, or retiring. This recipe uses agents to automate the tedious parts.

When to Use This

• Before a rebrand — Inventory everything that needs updating
• Quarterly content reviews — Keep documentation current
• SEO optimization — Find underperforming content to improve
• Knowledge base cleanup — Consolidate duplicate or outdated articles

The Pipeline

CRAWL     → Inventory all content assets
ANALYZE   → Evaluate each piece against criteria
RECOMMEND → Prioritized action plan

Step 1: Inventory

Create a database to store your content inventory:

@Assistant create a database called "Content Audit" 
with columns: title (text), url (text), type (text), 
last_updated (date), word_count (number), status (text), 
notes (text)

Then populate it:

@Researcher crawl our documentation at docs.example.com 
and add each page to the Content Audit database.
For each page, record the title, URL, content type 
(guide, reference, tutorial, FAQ), and approximate word count.

If you have content across multiple platforms, run separate crawls:

@Researcher also find all blog posts at example.com/blog 
and add them to the Content Audit database with type "blog"

Step 2: Analysis

With the inventory complete, analyze each piece:

@Analyst for each entry in the Content Audit database, 
visit the URL and evaluate:
- Is the content still accurate?
- Are there broken links?
- Is it well-structured (headings, lists, code examples)?
- Does it overlap with other content?
- Update the status column: "current", "needs-update", 
  "outdated", "duplicate", or "retire"
- Add specific notes about what needs fixing

Analysis Criteria

Customize the analysis based on your goals:

For SEO audits:

@Analyst also evaluate: keyword relevance,
meta description quality, internal linking,
and estimated search intent match


For documentation audits:

@Analyst also check: code examples work with current API version,
screenshots match current UI, terminology is consistent


Step 3: Recommendations

Generate an actionable report from the analysis:

@Writer query the Content Audit database for all entries 
where status is not "current" and create a document titled 
"Content Audit Recommendations" with:

1. Summary statistics (total pages, by status)
2. High-priority updates (outdated content with high traffic)
3. Content to consolidate (duplicates)
4. Content to retire (irrelevant or obsolete)
5. New content gaps (topics we should cover but don't)

Sort recommendations by priority and estimated effort.

Turning Recommendations into Tasks

Convert the recommendations into actionable work:

@Assistant for each recommendation in the Content Audit 
Recommendations document, create a task in the "Content Updates" 
task group with:
- Clear title describing the action
- Description with the specific changes needed
- Label: "high-priority", "medium-priority", or "low-priority"

Scheduling Regular Audits

For ongoing content health, set up a recurring audit using a scheduled skill:

1. Create an automation skill that runs monthly
2. The skill triggers the Researcher to check a sample of content
3. Results update the Content Audit database
4. A summary is posted to your #content channel

Variations

Documentation Version Audit

Check all docs against a new software version to find what needs updating.

Translation Audit

Compare content across languages to find untranslated or outdated translations.

Compliance Audit

Check all public-facing content against regulatory requirements or brand guidelines.