Content Refresh & Decay Agent
Identifies stale content, flags outdated statistics, and updates them — replacing old data with verified current sources, scored P0-P3 by severity.
What This Agent Does
Protects your existing content investment by systematically finding and fixing stale statistics, outdated references, broken links, and time-sensitive content that needs updating. Scans pages against your fact registry dates, flags deprecated tools and changed regulations, and outputs a prioritized refresh list scored from P0 (fix immediately) to P3 (next scheduled cycle).
Includes scheduled refresh cadences by content type and a fact registry update protocol that ensures traceability when statistics are replaced — logging old values, old sources, and dates so nothing is silently overwritten.
References & Sources
What You Get
- Staleness detection across 5 categories: statistics/data claims, technology/tool references, regulatory/compliance, competitive landscape, seasonal/time-sensitive content
- Priority scoring from P0 (critical — factually wrong) to P3 (low — minor date references)
- Refresh audit output with quoted stale content and specific recommended fixes
- Fact registry update protocol for traceability when updating statistics
- Scheduled refresh cadence recommendations by content type (homepage, blog, service pages, glossary)
Install
Choose your preferred installation method. Both put the agent rule in the right place for Claude Code to discover automatically.
Copy the rule below and save it as .claude/rules/content-refresh-decay.md in your project root.
# Content Refresh & Decay Agent Rules
When asked to run a content refresh audit, identify stale content, flag outdated statistics, and prioritize pages for refresh. This agent protects your existing content investment by ensuring content stays current.
## How to Run a Refresh Audit
1. Read the target page(s) completely
2. Cross-reference all statistics and data claims against your fact registry or data source dates
3. Check for outdated references (deprecated tools, changed regulations, old frameworks)
4. Scan for broken or likely-stale external links
5. Score staleness severity and output a refresh priority list
## Staleness Triggers
### Statistics & Data Claims
- **Flag:** Any statistic with a source date older than 18 months
- **Flag:** Any year reference that is 2+ years behind current year (e.g., "2024 data" when it's 2026)
- **Flag:** Market size figures older than 12 months (markets change fast)
- **Flag:** Any `[NEEDS VERIFICATION]` tags still in published content
- **Action:** Re-verify via your fact registry first, then web search if needed
### Technology & Tool References
- **Flag:** References to deprecated tools, APIs, or platforms
- **Flag:** Version numbers for tools/frameworks that have been superseded
- **Flag:** Algorithm update references that are no longer current (e.g., "Google's latest helpful content update" if a newer one exists)
- **Action:** Update to current version/name or remove if no longer relevant
### Regulatory & Compliance
- **Flag:** Regulatory references that may have changed (HIPAA updates, GDPR amendments, new AI regulations)
- **Flag:** Compliance frameworks with version changes (SOC 2, ISO updates)
- **Action:** Web search to verify current regulatory status before updating
### Competitive Landscape
- **Flag:** Competitor references where the company has been acquired, rebranded, or shut down
- **Flag:** Competitor feature/pricing claims older than 6 months
- **Action:** Verify current status, update or remove as needed
### Seasonal & Time-Sensitive Content
- **Flag:** "[Year] predictions" or "[Year] trends" posts that need year updates
- **Flag:** "This year" or "last year" references that are now stale
- **Flag:** Event references (conferences, launches) that have passed
- **Action:** Update year references or reframe as evergreen content
## Refresh Priority Scoring
| Factor | Weight | Scoring |
|---|---|---|
| Page traffic | High | High-traffic pages with stale stats = urgent |
| Staleness severity | High | Wrong data > outdated data > slightly old data |
| Page type | Medium | Hub pages > blog posts > glossary terms |
| Competitive impact | Medium | Pages ranking for competitive keywords need freshness |
| Number of stale elements | Low | More stale elements = higher priority (but one wrong stat is worse than five slightly old ones) |
### Priority Levels
| Priority | Criteria | Action Timeline |
|---|---|---|
| **P0 -- Critical** | Factually wrong stat, broken regulatory claim, deprecated tool presented as current | Fix immediately |
| **P1 -- High** | Stats >18 months old on high-traffic pages, year references 2+ years behind | Fix within 1 week |
| **P2 -- Medium** | Stats >12 months old on medium-traffic pages, slightly outdated tool versions | Fix within 1 month |
| **P3 -- Low** | Minor date references, low-traffic pages with old but not wrong data | Next scheduled refresh cycle |
## Refresh Rules
### What to Touch
- Update stale statistics with current, verified data
- Update year references and time-sensitive language
- Fix broken external links (replace with current URLs or remove)
- Update tool/technology references to current versions
- Refresh competitive claims with current data
### What NOT to Touch
- Page structure and heading hierarchy (unless specifically asked)
- Internal link architecture
- Voice and tone
- Schema and metadata structure
- Working content that isn't stale -- refresh is not rewrite
### Fact Registry Update Protocol
When updating a stat during refresh:
1. Log the old value, old source, and old date in a comment or changelog
2. Add the new value with new source and date
3. Never silently overwrite -- traceability matters
4. Format: `- [New stat claim] -- Source: [Name, Year](URL) (Updated from: [old value], [old source])`
## Refresh Audit Output Format
```
## Content Refresh Audit -- [Page URL or "All Blog Posts"]
**Scan date:** [Date] | **Pages scanned:** [N] | **Issues found:** [N]
### P0 -- Critical (Fix Immediately)
| Page | Issue | Current Content | Recommended Fix |
|---|---|---|---|
| /page | [description] | "[quoted stale content]" | [specific fix] |
### P1 -- High (Fix This Week)
| Page | Issue | Current Content | Recommended Fix |
|---|---|---|---|
### P2 -- Medium (Fix This Month)
| Page | Issue | Current Content | Recommended Fix |
|---|---|---|---|
### P3 -- Low (Next Refresh Cycle)
| Page | Issue | Current Content | Recommended Fix |
|---|---|---|---|
### Summary
- Pages scanned: [N]
- P0 issues: [N] | P1: [N] | P2: [N] | P3: [N]
- Fact registry entries to update: [N]
- Estimated refresh effort: [Small/Medium/Large]
```
## Scheduled Refresh Cadence
| Content Type | Refresh Cycle | What to Check |
|---|---|---|
| Blog posts | Every 6 months | Stats, links, year references, relevance |
| Service/hub pages | Every 3 months | Methodology currency, competitive claims, CTA relevance |
| Vertical/industry pages | Every 6 months | Regulatory changes, vertical-specific data, competitor landscape |
| Glossary/reference terms | Every 12 months | Definition accuracy, related terms currency |
| Homepage | Every 3 months | Value prop currency, featured stats, social proof |
---
## Out of Scope
This agent focuses on identifying and prioritizing stale content. The following are not covered:
- Full content rewrites (flag issues, hand off to content writers)
- Internal link architecture changes
- Schema and metadata structural updates
- New content creation and editorial planning
- Visual design or layout updatesUsage
Once installed, open your project in Claude Code and ask:
Run a content refresh audit on all blog posts and update any stale statisticsClaude Code will follow the scoring rubric, check every dimension, and output a structured scorecard with pass/fail per check and prioritized fix recommendations.
Works Great With
Fact Verification Agent
Enforces a verification workflow for all statistics and data claims — checks your fact registry, searches for sources, and rewrites unverifiable claims with qualitative language.
Content QA Audit Agent
68 checks across 7 dimensions: Technical SEO, Linking, E-E-A-T, Anti-Slop, AI Extraction, Schema, and AI Access. Scores your page and fixes what fails.
Copywriter Audit Agent
40 checks across 4 dimensions: Audience Resonance, Authority Positioning, Trust & Authenticity, Anti-Formulaic Detection. Scores your copy and rewrites what falls short.
Need a Custom Agent?
We build custom Claude Code agent rules tailored to your team's workflows, content standards, and tech stack.
Get in touch