SEO Intelligence¶

Umbrella for 7 capabilities — see the cluster table below for slugs, costs, and links Plan Tier: Accelerate (most capabilities) / Command (full set) Backends: DataForSEO (4 capabilities) + Claude Sonnet (3 capabilities)

Overview¶

SEO Intelligence is the umbrella for the seven agent capabilities that cover an end-to-end SEO motion — from keyword discovery and competitive analysis through to content optimization and meta-tag generation. The four capabilities that need live search data call DataForSEO. The three capabilities that need structured rewriting or scoring call Claude. Each capability is independently scoped — partners can include only the ones they need on a Partner API key.

This page is the consolidated reference for the whole cluster. Each capability has its own slug, cost, and parameter shape, but the workflow they enable runs naturally end-to-end.

The seven capabilities at a glance¶

#	Slug	Name	Backend	Credit cost
1	`seo_intelligence`	SEO Intelligence (domain analysis)	DataForSEO	`count(analyses) × 1` — default 3
2	`keyword_research`	Keyword Research	DataForSEO	5 (flat)
3	`content_gap_analysis`	Content Gap Analysis	DataForSEO	`max(2, competitor_count + 1) × 2` — minimum 4
4	`discover_search_questions`	Discover Search Questions (PAA)	DataForSEO	3 (flat)
5	`seo_content_optimize`	SEO Content Optimization	Claude	12 (flat)
6	`seo_content_analysis`	SEO Content Analysis	Claude	8 (flat)
7	`generate_meta_tags`	Generate SEO Meta Tags	Claude	5 (flat)

Per-capability detail follows.

1. `seo_intelligence` — domain analysis¶

The headline capability. Runs one or more DataForSEO domain-analytics queries against the target domain and returns a consolidated report.

Required: domain. Default analyses: ['overview', 'competitors', 'keywords'] — selectable from this set:

Analysis	What you get
`overview`	Domain rank, organic traffic estimate, keyword count, top categories
`competitors`	Competitor domains and overlap signals
`keywords`	Ranked keywords with positions, search volumes, difficulty
`technologies`	Technology stack detected on the domain
`whois`	Registration data and domain history

Credit cost: count(analyses) × 1. Three analyses = 3 credits. All five = 5 credits. The DataForSEO call rate is API_DOMAIN_ANALYTICS = 1 per query in App\Models\SeoUsageLog::DEFAULT_CREDIT_COSTS.

Optional parameters: location (defaults to United States).

Returns: Structured per-analysis results plus an analyses_run array confirming which analyses produced data, and credits_used for billing transparency.

2. `keyword_research` — search-volume discovery¶

Pulls real-time search volumes, keyword difficulty, CPC data, and related-keyword expansions from DataForSEO. Used for picking target keywords before writing content.

Required: topic. Cost: flat 5 credits.

Returns: Search volumes (monthly), keyword difficulty (0-100), CPC data for paid-comparison purposes, related-keyword tree, and long-tail variations.

When to use: as the first step in any SEO motion, before committing to topic angles.

3. `content_gap_analysis` — competitor keyword gaps¶

Identifies keywords your competitors rank for that you don't. Useful for finding content territory to claim.

Required: domain. Optional: competitors (array of competitor domains). Cost: max(2, count(competitors) + 1) × 2 credits. Minimum cost is 4 credits (2 × API_LABS rate of 2). Five competitors = 6 × 2 = 12 credits.

Returns: Keywords ranked for by competitors but not by your domain, with search volume, difficulty, and an total_volume metric that sums opportunity size.

The 2-credit-per-call rate comes from API_LABS in SeoUsageLog — DataForSEO's "Labs" endpoints cost more than basic domain analytics because the data is more derived.

4. `discover_search_questions` — Google People Also Ask¶

Extracts the "People Also Ask" questions from Google search results for a topic, plus search-intent classification. Useful for FAQ content, blog ideation, and learning what your audience actually asks.

Required: topic. Cost: flat 3 credits.

Returns: PAA questions list, related question variations, search-intent classification per question (informational / navigational / commercial / transactional), and frequency hints.

5. `seo_content_optimize` — AI rewrite for keywords¶

Takes existing content and a target keyword; returns a rewrite optimized for that keyword with related-term coverage. Claude-backed.

Required: content, keyword. Cost: flat 12 credits.

Returns: The rewritten content, a list of keywords woven in, and a delta summary noting what changed and why.

When to use: after keyword_research identifies a target keyword you want a piece of content to own.

6. `seo_content_analysis` — 17-factor on-page scoring¶

Scores a piece of content against 17 SEO factors and returns per-factor pass/fail with actionable fixes. Claude-backed.

Required: content, keyword. Optional: content_type (default article). Cost: flat 8 credits.

Returns: Overall score (0-100), per-factor breakdown, list of actionable fixes ordered by impact, and a "would publish at this score" recommendation. The 17 factors cover keyword usage, heading structure, link density, readability, and so on.

When to use: as a final pre-publish QA pass, after seo_content_optimize if you've rewritten the content.

7. `generate_meta_tags` — titles and descriptions¶

Generates SEO-optimized meta titles and descriptions for a target keyword. Claude-backed.

Required: keyword. Optional: meta_type (default all — title + description; can also request just one). Cost: flat 5 credits.

Returns: Suggested title (under 60 characters), suggested description (under 160 characters), and 2-3 alternates for each so you can pick the one that fits your voice.

When to use: just before publishing — close to last step in the SEO motion.

A typical end-to-end SEO motion¶

The capabilities chain naturally:

keyword_research               → "What keywords matter?"
   ↓
seo_intelligence               → "How well does my domain cover them?"
(overview + keywords)
   ↓
content_gap_analysis           → "Which keywords do competitors own that I don't?"
   ↓
discover_search_questions      → "What do users actually ask about this topic?"
   ↓
[ write or commission content ]
   ↓
seo_content_optimize           → "Rewrite the draft to target the keyword"
   ↓
seo_content_analysis           → "Score the draft and surface fixes"
   ↓
generate_meta_tags             → "Title and description before publishing"

Total credits for one full pass with default settings: roughly 5 + 3 + 4 + 3 + 12 + 8 + 5 = 40 credits for a complete keyword → published-content motion. Less if you skip steps; more if your competitor list is large for the gap analysis.

Credit costs reference¶

Sourced from getEstimatedCredits() on each capability handler. Base call rates from App\Models\SeoUsageLog::DEFAULT_CREDIT_COSTS.

Capability	Cost formula	Default
`seo_intelligence`	`count(analyses) × 1`	3 (overview + competitors + keywords)
`keyword_research`	flat	5
`content_gap_analysis`	`max(2, competitor_count + 1) × 2`	4 (no competitors) — 12 (5 competitors)
`discover_search_questions`	flat	3
`seo_content_optimize`	flat	12
`seo_content_analysis`	flat	8
`generate_meta_tags`	flat	5

DataForSEO endpoint rates (per call):

Endpoint family	Rate	Used by
`API_DOMAIN_ANALYTICS`	1 credit	`seo_intelligence` (each analysis = 1 call)
`API_LABS`	2 credits	`content_gap_analysis`

Constants are tunable in SeoUsageLog::DEFAULT_CREDIT_COSTS — update them centrally when DataForSEO pricing changes.

Use cases¶

Goal	Capabilities involved
Pick keywords for a new blog post	`keyword_research` + `discover_search_questions`
Audit a domain's SEO position	`seo_intelligence` (overview + competitors + keywords)
Find content opportunities	`content_gap_analysis`
Optimize existing content for a target keyword	`seo_content_optimize` + `seo_content_analysis`
Generate listing titles and descriptions	`generate_meta_tags`
Pre-publish quality gate	`seo_content_analysis` + `generate_meta_tags`

What this is not¶

Agent Discoverability Score. The marketplace_listing_seo capability tracks AI-assistant mentions (ChatGPT, Claude, Perplexity) and backlinks specifically for a marketplace listing. Different concern, different capability — see the Agent Discoverability Score page.
A scraper. All search data is sourced from DataForSEO's official API; the Claude-backed capabilities operate on content you provide.
A rank tracker. The capabilities surface point-in-time rankings, not historical trend data.

Plan availability¶

Capability	Starter	Accelerate	Command
`seo_intelligence`	—	✅	✅
`keyword_research`	—	✅	✅
`content_gap_analysis`	—	✅	✅
`discover_search_questions`	—	✅	✅
`seo_content_optimize`	—	✅	✅
`seo_content_analysis`	—	✅	✅
`generate_meta_tags`	—	✅	✅
`marketplace_listing_seo` (separate)	—	—	✅

API access (api_access scope) starts on Accelerate — required for any programmatic call regardless of capability. See Partner API Capabilities for the full scoping reference.

API access¶

All seven capabilities are exposed through the agent runtime under their snake_case slugs. Partner API keys scoped with manage_seo plus the agent capability slugs you need can run them programmatically.

Sample scoping for a content team:

"scoped_capabilities": [
  "api_access",
  "manage_seo",
  "manage_content",
  "keyword_research",
  "discover_search_questions",
  "seo_content_optimize",
  "seo_content_analysis",
  "generate_meta_tags"
]

See Partner API Capabilities and Authentication & API Keys.

SEO Intelligence¶

Overview¶

The seven capabilities at a glance¶

1. seo_intelligence — domain analysis¶

2. keyword_research — search-volume discovery¶

3. content_gap_analysis — competitor keyword gaps¶

4. discover_search_questions — Google People Also Ask¶

5. seo_content_optimize — AI rewrite for keywords¶

6. seo_content_analysis — 17-factor on-page scoring¶

7. generate_meta_tags — titles and descriptions¶