Loading ToolBench
MCP server para obter informações de uso do GitHub Copilot
This server has critical definition quality gaps. All three tools have NO input parameters (empty schemas with no properties), which is appropriate for read-only discovery tools but violates the pattern of including parameter descriptions where applicable. More critically, descriptions are provided in Portuguese and are somewhat generic. Tool names follow verb_noun convention (get_*) which is positive, but lack of structured output documentation, no error recovery guidance, and absence of parameter validation details are major issues. The server appears to be a simple read-only API wrapper with minimal sophistication, suitable for its use case but falling well below production-grade tool patterns.
Obtém informações de uso atual do GitHub Copilot, incluindo cotas e limites, dados originais da API
Obtém informações de uso do GitHub Copilot formatado humanizado
Obtém um resumo conciso do uso do GitHub Copilot com informações principais, como o restante da quota premium (economiza tokens)
All tool descriptions are in Portuguese, making them inaccessible to English-speaking LLM agents. Descriptions should be in English for international compatibility.
No output schema documented for any tool. LLMs cannot determine what fields to expect in responses, making downstream planning and field extraction error-prone. Documentation must specify: for get_copilot_usage (raw API response structure), for get_copilot_usage_formatted (text string), for get_copilot_usage_summary (text string with summary format).
No error recovery guidance in tool descriptions. The callToolHandler catches errors and returns an error response, but descriptions do not explain: What can go wrong? (missing token, network error, invalid credentials). What should the agent do? Descriptions should state: 'Requires COPILOT_TOKEN environment variable. Returns error if token is invalid or GitHub API is unreachable.'
| Scored | Grade | Overall | Spec posture | Rubric |
|---|---|---|---|---|
| 2026-09-22 | F | 11 | 2026-07-28+ | v2 |
| 2026-03-09 | F | 33 | - | v1 |
Descriptions lack WHEN to use and WHAT distinguishes each tool. 'get_copilot_usage_formatted' description ('Obtém informações de uso do GitHub Copilot formatado humanizado') does not explain: when should agent choose formatted vs raw vs summary? All three descriptions are vague about trade-offs.
All three tools have empty input schemas (properties: {}, required: []). While this is correct for parameter-less tools, it means no parameter descriptions exist. Baseline expectation: even zero-parameter tools should briefly document prerequisites (e.g., 'Requires COPILOT_TOKEN environment variable to be set').