Loading ToolBench
moz-mcp provides 13 tools with consistent naming patterns and mostly clear descriptions. However, there are significant gaps in schema completeness, parameter documentation, and output schema definition. Tools follow verb_noun naming conventions (moz_keyword_*, moz_site_*) which aids clarity, but descriptions lack depth about WHEN to use specific tools vs alternatives, and most tools lack documented output schemas. All 13 tools have descriptions (90-150 chars typically), but parameter descriptions are minimal and output structures are completely undocumented. No error handling guidance is visible in the tool definitions. Security is well-handled via environment variable injection for API credentials, but this is baseline practice, not exceptional.
Comprehensive competitor analysis by combining site metrics, ranking keywords, and providing competitor identification guidance. Note: Moz API does not automatically identify competitors, but this tool helps analyze potential competitors you specify.
Fetch keyword difficulty score
Fetch all keyword metrics (difficulty, volume, organic CTR, priority)
Fetch keyword opportunity (organic CTR) data
Fetch keyword priority score
Fetch search intent data for a keyword
No output schemas documented for any tool. LLMs cannot determine what fields to expect, which breaks downstream tool chaining and forces agents to guess field names.
Parameter descriptions are minimal or missing usage context. E.g., 'locale (defaults to en-US if not specified)' lacks format guidance (e.g., is 'en-US' or 'en_US' the valid format?). 'engine (google, bing)' is vague, should it be lowercase? What happens if an unsupported engine is passed?
| Scored | Grade | Overall | Spec posture | Rubric |
|---|---|---|---|---|
| 2026-09-22 | D | 51 | 2026-07-28+ | v2 |
| 2026-03-09 | F | 0 | - | v1 |
Get related keyword suggestions
Fetch keyword search volume
Check your Moz API quota and usage limits
Fetch Brand Authority for a site
Fetch metrics for a single site
Fetch metrics for multiple sites at once
List ranking keywords for a site
No enum constraints for 'engine' parameter across multiple keyword tools. Tools accept 'google, bing' but parameter lacks explicit enum definition. LLMs may hallucinate invalid values like 'duckduckgo', 'yandex'.
No pagination support documented or apparent for tools returning lists (moz_keyword_suggestions, moz_site_ranking_keywords, moz_competitor_analysis). Default limits may cause truncation, and no offset/limit parameters are visible in schemas.
No error handling guidance in tool descriptions. What happens if a keyword is invalid, or a site domain cannot be reached? How should an agent recover? No recovery instructions documented.
Tool descriptions do not clarify WHEN to use overlapping tools. E.g., when to use moz_keyword_difficulty vs moz_keyword_metrics vs moz_keyword_priority is unclear. Agents may call all three when one suffices.
moz_competitor_analysis description notes that 'Moz API does not automatically identify competitors' but the tool still requires 'competitor_sites' as a required input. This is unclear, should competitor_sites be optional? The description should better explain the workflow.
No tool annotations (readOnlyHint, destructiveHint, idempotentHint) are present. All tools appear to be READ_ONLY based on risk field, but this is not declared in schemas for agent planning.