Loading ToolBench

Reading the current corpus…

k6-mcps-tests-http-azdevops-pipelines MCP server · F · ToolBench

k6-mcps-tests-http-azdevops-pipelines

github.com/renatogroffe/k6-mcps-tests-http-azdevops-pipelines

Load testing suite using k6 to test MCP (Model Context Protocol) HTTP servers with currency quotation tools, designed for Azure DevOps Pipelines integration

2026-07-28+STREAMABLE_HTTP3 tools

31/100

Definition Quality35

Protocol Readiness15

Supportability36

Spec posture

Inferred effective spec 2026-07-28+·current baseline 2026-07-28

Static source inference · medium confidence · evidence: Streamable HTTP

Current-spec patterns detected

Summary

This server has three tools with minimal definition quality. All tools have basic descriptions (40-60 chars, within acceptable range) and identical input schemas (numberOfRecords: integer). However, parameter descriptions are missing entirely, the schema shows only type and no description field. Tool names use Portuguese verbs (obter_*) which is non-standard for English-speaking LLM agents. No output schemas are documented. No error handling guidance. The server appears to be a k6 load-testing harness rather than a production MCP server, it tests tools via HTTP calls but does not implement the MCP protocol itself (no tool registration, no schema validation, no error recovery patterns).

Tools (3)

obter_cotacao_dolarread onlysource verified35/100

Obtains the current USD (Dollar) exchange rate quotation

obter_cotacao_euroread onlysource verified35/100

Obtains the current EUR (Euro) exchange rate quotation

obter_cotacao_libraread onlysource verified35/100

Obtains the current GBP (British Pound) exchange rate quotation

Issues (5)

CRITICALtool-description ↗

Parameter descriptions missing. Schema shows only type (integer) with no description field explaining what numberOfRecords controls, valid range, or constraints.

obter_cotacao_dolarobter_cotacao_euroobter_cotacao_libra

CRITICALresponse-shaper ↗

No output schema documented. Tools return JSON with fields like isSuccess, data.moeda, data.valor, but LLMs cannot infer structure. Must document return type explicitly.

obter_cotacao_dolarobter_cotacao_euroobter_cotacao_libra

HIGHtool ↗

Tool names in Portuguese (obter_*) are non-standard. English-speaking LLM agents expect verb_noun in English (e.g., get_dollar_rate). This creates friction for multi-lingual agent deployments.

obter_cotacao_dolarobter_cotacao_euroobter_cotacao_libra

HIGHrecovery-guide ↗

No error handling guidance. If numberOfRecords is negative, zero, or exceeds API limits, no recovery instructions provided. LLM cannot self-correct.

Recommendations

Add parameter descriptions to numberOfRecords: 'Number of exchange rate records to retrieve (1-100, default 10). Higher values return historical data.'
Document output schema: {isSuccess: boolean, message: string, data: {moeda: string, valor: number, timestamp: ISO8601}}
Rename tools to English: get_dollar_rate, get_euro_rate, get_pound_rate. Use verb_noun convention matching 549-tool baseline.
Add numeric constraints: numberOfRecords minimum 1, maximum 100, with clear error message if violated.
Implement actual MCP server (not k6 test harness). Register tools via MCP protocol with proper schema validation, error recovery, and stateless request handling per spec 2026-07-28.

Score history

Overall score trend

First recorded score · v2 rubric

31/100

Scored	Grade	Overall	Spec posture	Rubric
2026-09-23	F	31	2026-07-28+	v2