MCP Tools Reference

unihra_health

Check API availability.

Input: (none)

Output: {"status": "ok", "version": "..."}

---

unihra_get_limits

Show API key usage limits and balance.

Input: (none)

Output:

{
  "plan": "pro",
  "daily_limit": 5000,
  "used_today": 12,
  "remaining_today": 4988
}

---

unihra_analyze

Run SEO analysis. Primary tool — the model should call this first for any page comparison request.

Input	Type	Required	Description
own_page	string	✅	Your page URL
competitors	string[]	✅	Competitor URLs (3–10 recommended)
queries	string[]	—	Target search queries (up to 5)
lang	"ru" | "en"	—	Page language
url_cookies	object	—	Per-URL auth cookies
triplet_analysis	boolean	—	Enable Knowledge Graph (5 credits, default false)

Output: Compact summary with result_id + top-5 gaps, action counts, missing n-grams, and structure snapshot.

---

unihra_list_results

List all saved analysis results in the current session directory.

Input: (none)

Output: {"count": N, "results": [{"result_id": "...", "own_page": "...", "saved_at": "...", ...}]}

---

unihra_delete_result

Delete a saved result from disk.

Input	Type	Required	Description
result_id	string	✅	From unihra_analyze or unihra_list_results

---

unihra_get_page_status

Get requested URL coverage for a saved analysis. The tool returns only public page lists: URLs included in the analysis and URLs not included.

Input	Type	Required	Description
result_id	string	✅	From unihra_analyze

Output:

{
  "result_id": "abc123",
  "parsed_count": 1,
  "failed_count": 1,
  "parsed": [{"url": "https://example.com", "is_own_page": true}],
  "failed": [{"url": "https://competitor.example", "is_own_page": false}]
}

---

unihra_get_gaps

Get Umbrella Analysis gaps — words to add to your page.

Input	Type	Default	Description
result_id	string	required	—
top_n	integer	50	Max items to return
min_coverage	number	15.0	Min competitor coverage %
min_gap	number	0.0	Min gap score (0 = show all)
missing_only	boolean	false	Only words absent from own page
group_by_recommendation	boolean	false	Group by placement recommendation

---

unihra_get_anchors

Get anchor text analysis.

Input	Type	Default	Description
result_id	string	required	—
top_n	integer	50	Max items
missing_only	boolean	false	Only anchors absent from own page

---

unihra_get_word_actions

Get TF-IDF word actions (add / increase / decrease / ok).

Input	Type	Default	Description
result_id	string	required	—
action	"add" | "increase" | "decrease" | "ok" | "all"	"all"	Filter by action
top_n	integer	50	Max items per group
min_frequency	number	1.0	Min word frequency
exclude_ok	boolean	true	Exclude ok-action words

---

unihra_get_ngrams

Get phrase (n-gram) analysis.

Input	Type	Default	Description
result_id	string	required	—
top_n	integer	60	Max phrases
min_pages_count	integer	2	Min competitor sites using this phrase
missing_only	boolean	false	Only phrases not on own page
ngram_type	"all" | "bigrams" | "lemma_trigrams"	"all"	Filter by phrase length

---

unihra_get_triplets

Get Knowledge Graph entities and topical gaps. Only available for results created with triplet_analysis=true.

Input	Type	Default	Description
result_id	string	required	—
view	"entities" | "gaps" | "both"	"both"	Which view to return
top_entities	integer	30	Max entities
max_triplets_per_entity	integer	12	Max facts per entity
tiers	string[]	["core","main"]	Filter entities by tier
gap_severities	string[]	["critical","important"]	Gap severity buckets to include
max_gaps_per_severity	integer	25	Max gap items per bucket

---

unihra_get_page_structure

Get heading structure and meta tags.

Input	Type	Default	Description
result_id	string	required	—
own_only	boolean	false	Return only your page structure