Academic Writer MCP

Tester un fournisseur d'IA

test_ai_provider

Read-only

Verifies an AI provider (Anthropic, OpenAI, or Ollama) with a minimal Reply OK request and no project data, confirming connectivity and local/cloud token behavior before writing.

Instructions

Vérifie un moteur par un très court appel (« Réponds OK »), sans aucune donnée de projet. anthropic/openai : consomme quelques tokens et nécessite le mode connecté ; ollama : local et gratuit.

Tool call template

This tool declares no required parameters. Add optional parameters from the input schema as needed.

{
  "id": 1,
  "jsonrpc": "2.0",
  "method": "tools/call",
  "params": {
    "arguments": {},
    "name": "test_ai_provider"
  }
}

Input Schema

Name	Required	Description	Default
`provider`	No

Output Schema

Name	Required	Description
`text`	Yes
`model`	Yes
`provider`	Yes
`served_by`	No	Modèle ayant réellement répondu (repli éventuel).
`truncated`	No	Réponse coupée par la limite de tokens.
`request_id`	No
`stop_reason`	No
`input_tokens`	No
`output_tokens`	No

Schema Changelog

Changes observed during successful MCP inspections.

First observedv1.0.0Nov 9, 2026

TDQS

A3.9/5.0

Behavior4/5

Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

Annotations only declare readOnlyHint/openWorldHint/idempotentHint; the description adds substantive context beyond them: a minimal prompt is sent, no project data is transmitted, anthropic/openai consume a few tokens and require connected mode, ollama is local and free. It does not say what a failed test looks like or whether results are cached, but the cost/connectivity disclosure is real added value.

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Conciseness5/5

Is the description appropriately sized, front-loaded, and free of redundancy?

Two dense lines: the core action and its no-project-data guarantee come first, then the per-provider cost/connectivity note. The quoted prompt ("Réponds OK") and the provider split each earn their place; nothing is padded.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Completeness4/5

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

An output schema exists, so return values need not be explained, and the description covers scope, cost, and connectivity prerequisites for a one-parameter tool. Minor gaps remain: no mention of failure behavior or whether connected mode is a precondition the caller must satisfy beforehand.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Parameters4/5

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

Schema description coverage is 0% and the single optional parameter is an enum, so the description carries the burden; it explains what each provider choice costs (tokens + connected mode vs. local and free) and that omitting it is acceptable. That is meaningfully more than the bare enum in the schema.

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Purpose4/5

Does the description clearly state what the tool does and how it differs from similar tools?

States a specific verb and resource: it verifies an AI engine via a very short call ("Réponds OK") and explicitly scopes it as using no project data, which implicitly separates it from project-analysis siblings like detect_ai_indicators. It stops short of naming a sibling such as get_ai_status, so an agent must infer which one answers "is the AI working?".

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Usage Guidelines3/5

Does the description explain when to use this tool, when not to, or what alternatives exist?

Implies usage (a connectivity/smoke check before real work) through "très court appel" and "sans aucune donnée de projet", and gives per-provider conditions (connected mode required for anthropic/openai, ollama is local and free). It never states explicitly when to choose this over get_ai_status or when not to call it, leaving routing to inference.

Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.

Deploy Server