Loading ToolBench

Reading the current corpus…

mcp-google-trends MCP server · D · ToolBench

mcp-google-trends

Free access to Google Trends data. Provides 5 tools for search trend analysis: interest over time, keyword comparison, related queries, trending searches, and regional interest. No API keys required - uses trendspy library (unofficial Google Trends).

<=2025-11-25STDIO5 tools

54/100

Definition Quality74

Protocol Readiness47

Supportability25

Spec posture

Inferred effective spec <=2025-11-25·current baseline 2026-07-28

Static source inference · medium confidence · detected: Logging

Deprecated protocol patterns detected

Summary

Server provides 5 tools with complete input schemas and descriptions. All tools follow verb_noun naming (get_*, compare_*). Descriptions are present and adequate (100-150 chars avg), exceeding the 20-char minimum. However, output schemas are not formally documented, responses are inferred from code. Parameter descriptions lack constraint details (e.g., 'geo' accepts region codes but no format guidance; 'timeframe' accepts strings like 'today 12-m' but no enum or pattern). Error handling is minimal, no recovery guidance or actionable error messages. No tool annotations (readOnlyHint, etc.). Caching logic is internal and not exposed to the agent. Overall: solid naming and basic descriptions, but missing schema documentation, constraint clarity, and error guidance expected of production tools.

Tools (5)

compare_keywordsread onlysource verified75/100

Compare search interest for multiple keywords

get_interest_by_regionread onlysource verified72/100

Get search interest broken down by geographic regions

get_interest_over_timeread onlysource verified75/100

Get search interest over time for a keyword

get_related_queriesread onlysource verified75/100

Get related queries and search suggestions for a keyword

get_trending_searchesread onlysource verified73/100

Get currently trending searches in a geographic region

Issues (5)

HIGHtool ↗

Output schemas not formally documented. Responses are inferred from code (e.g., _get_interest_over_time returns dict with 'data', 'summary', 'message' fields), but no schema is exposed to the agent. LLMs cannot plan downstream operations without knowing response structure.

get_interest_over_timecompare_keywordsget_related_queriesget_trending_searchesget_interest_by_region

HIGHconstrained-input ↗

Parameter constraints not specified. 'timeframe' accepts strings like 'today 12-m' but no enum or regex pattern is documented. 'geo' accepts region codes but format is not explained. 'resolution' in get_interest_by_region lists enum values in description ('COUNTRY, REGION, CITY') but not as a formal enum constraint.

get_interest_over_timecompare_keywordsget_related_queriesget_interest_by_region

HIGHrecovery-guide ↗

Error handling lacks recovery guidance. When no data is available, tools return a 'message' field but do not suggest next steps (e.g., 'Try a different timeframe or geo region'). Errors are not categorized as retryable, user-fixable, or fatal.

Recommendations

Document output schemas for all tools. Add a 'returns' section to each tool definition specifying field names, types, and descriptions. Example: 'Returns {data: [{date: string, value: int}], summary: {max: int, min: int, avg: float, current: int, trend_direction: string}}'.
Convert parameter constraints to formal enums or patterns. For 'timeframe', define enum: ['today 1-m', 'today 3-m', 'today 12-m', 'today 5-y', 'all']. For 'resolution', use enum: ['COUNTRY', 'REGION', 'CITY']. For 'geo', document the format (e.g., 'ISO 3166-1 alpha-2 country code or region code').
Add recovery guidance to error responses. When no data is available, return: 'No data available for keyword="{keyword}" in timeframe="{timeframe}". Try: (1) expanding timeframe to 'today 5-y', (2) using a broader keyword, or (3) checking geo region availability.'
Add tool annotations. Declare all tools with readOnlyHint=true since they do not modify state. This signals to agents that retries are safe.
Implement pagination for list-like tools. Add 'offset' and 'limit' parameters to get_trending_searches and get_interest_by_region. Return 'total_count' and 'has_more' in responses to enable multi-page iteration.

Spec posture evidence

Inferred effective spec: <=2025-11-25.

Relies on Logging (deprecated) - log to stderr or use OpenTelemetry

Score history

Overall score trend

First recorded score · v2 rubric

54/100

Scored	Grade	Overall	Spec posture	Rubric
2026-09-23	D	54	<=2025-11-25	v2