Content Strategy Agent: build topics from public demand, not guesses
How the UnifAPI Content Strategy Agent combines SEO, Reddit, YouTube, TikTok, Threads, X, and news records into sourced content opportunities.
In short: The Content Strategy Agent turns public demand signals into a prioritized content backlog. It reads search demand, community questions, video trends, social language, and news context before it recommends topics.
Most content plans start with a brainstorm. That can work, but it is fragile. The better starting point is public demand: what people search, ask, complain about, watch, share, and cite.
The Content Strategy Agent uses UnifAPI records to build that evidence base inside Claude or Codex.
What it reads
| Source | Signal |
|---|---|
| SEO | Keyword ideas, related queries, overview metrics, history, intent |
| Community questions, objections, language, thread depth | |
| YouTube | Videos, channels, related content, topic momentum |
| TikTok | Short-form trend and creator signals |
| X and Threads | Real-time market language and founder/customer discussion |
| News | Market context and recent events |
The Agent does not need every source for every run. A good Skill scopes the research to the decision.
A copyable run prompt
Build a content opportunity brief for example.com, a public-data API for marketing Agents. Use UnifAPI MCP to research search demand, Reddit questions, YouTube or TikTok topic momentum, and recent news. Return 12 content opportunities grouped by buyer stage. For each opportunity, include audience, search/query evidence, community language, suggested angle, difficulty or confidence, and internal link target. Separate observed evidence from recommendations.
This asks for a content backlog, not a generic article list. The output should help a team decide what to write and why.
The output format
| Field | Why it matters |
|---|---|
| Topic | The concrete article or page idea |
| Buyer stage | Awareness, consideration, decision, or implementation |
| Evidence | Search, community, social, video, or news records |
| Angle | The specific claim or framing |
| Fit | Why the topic matches the product |
| Confidence | How strong the evidence is |
| Next asset | Blog post, guide, comparison page, Skill page, or docs page |
The "next asset" field prevents a common mistake: sending every idea to the blog. Some topics belong on /agents, some on /skills, some in docs, some as comparison pages, and some as blog posts.
Where Skills help
The Content Strategy Agent can use several Skills:
| Skill | Job |
|---|---|
content-opportunity-brief | Turn public demand into a prioritized content backlog |
content-strategy | Decide pillars, topic clusters, and sequencing |
customer-research | Mine public language and objections |
| Platform Data Skills | Read Reddit, YouTube, TikTok, X, Threads, or news deterministically |
The Skill layer keeps the assistant from doing shallow "10 blog ideas" output. It forces the run to include evidence and prioritization.
A good topic has two kinds of fit
Search fit means people are looking for it. Product fit means UnifAPI has a credible reason to answer it.
| Topic | Search fit | Product fit |
|---|---|---|
| "SEO Agent for Claude" | High-intent long-tail | Strong: SEO Agent + MCP |
| "How to make a viral TikTok" | Broad demand | Weak unless tied to public trend research |
| "Reddit research agent" | Clear source intent | Strong: Reddit Data Skill + Social Listening Agent |
| "Marketing automation software" | Crowded head term | Weak if it implies write actions |
This is where the Agent should be conservative. Do not chase every keyword. Chase the ones where the product can provide a better answer.
What to avoid
The Content Strategy Agent should not invent volume, fabricate quotes, or treat one social post as market proof. It should not recommend content that claims UnifAPI posts, edits, or executes actions on behalf of the user.
It should also avoid doorway pages. If a vertical or platform page is recommended, it needs genuine industry or source substance, not just a swapped noun.
What to read next
Platform Agents and Data Skills - explains the source-specific read paths the Content Strategy Agent uses.
Vertical marketing Agents - shows how content strategy changes by industry.
Browse the Content Strategy Agent - open the live Agent page.