Generates constructive grading feedback and suggested scores from a question, maximum score, student answer, and optional reference answer, helping teaching assistants reduce manual review effort.
【AI 智能批改輔助】提供題目、滿分、學生作答內容與參考擬答，AI 自動給予具體建設性的批改評語草稿（優點肯定、盲點指正、未來精進建議）與建議給分，大幅減輕助教人工批改負擔！
| Name | Required | Description | Default |
|---|---|---|---|
| maxScore | Yes | 該題滿分（例如 30） | |
| studentAnswer | Yes | 學生提交之作答內容 | |
| questionPrompt | Yes | 題目內文與題旨 | |
| referenceAnswer | No | 參考擬答或關鍵評分規準 |
Changes observed during successful MCP inspections.
v1.0.0Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
With no annotations, the description carries the full burden. It does add real value by disclosing the output structure (three feedback dimensions plus a suggested score) and by framing the result as a '草稿' (draft), implying nothing is committed. However, it never states whether the suggested score is persisted, whether the call mutates any submission record, or anything about latency, cost, or the reliability of AI-generated judgement.
Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.
Is the description appropriately sized, front-loaded, and free of redundancy?
A single compact sentence bracketed by a tag, front-loading both the input contract and the deliverables. Every clause earns its place; the only minor cost is the marketing-flavored tail about reducing TA workload, which adds tone rather than information.
Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.
Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?
There is no output schema, and the description compensates by enumerating what the response contains (three-part feedback and a suggested score), which is exactly what an agent needs to chain it with a grading write step. Combined with fully documented parameters, the definition is nearly complete; only the side-effect question (does it store anything?) is left open.
Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.
Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?
Schema description coverage is 100%, with each of the four properties documented, so the baseline is 3. The description names the same four inputs (題目、滿分、學生作答內容與參考擬答), effectively restating the schema rather than adding semantics such as how referenceAnswer influences scoring or what happens when it is omitted (it is the only non-required parameter).
Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.
Does the description clearly state what the tool does and how it differs from similar tools?
States a specific verb and resource: given question, max score, student answer and reference answer, the AI produces a constructive grading-comment draft (praise, blind spots, improvement suggestions) plus a suggested score. The operation and its output shape are unmistakable. It does not distinguish itself from nearby siblings such as audit_essay_question or ai_generate_homework_assignment, so it falls short of a 5.
Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.
Does the description explain when to use this tool, when not to, or what alternatives exist?
The phrase '大幅減輕助教人工批改負擔' implies the TA-grading context, so when to use it is inferable, but there is no explicit condition, prerequisite, or named alternative (e.g. calling audit_essay_question first, or whether a submission must already exist). Guidance exists only by implication.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.