build_summary_prompt

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

Annotations already declare readOnlyHint=true, idempotentHint=true, destructiveHint=false. The description goes beyond this by disclosing the key behavioral trait: the MCP makes no LLM API call and is a pure assembly operation. Knowing this tool has zero external side effects and does not itself perform the summary is genuinely valuable context that the annotations do not fully convey. No contradiction with the annotations; the description reinforces them.

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?

The purpose is front-loaded in the first sentence, followed by a valuable clarifier about not calling the LLM, then clean Args/Returns sections. The structure is tidy and scannable. The Args section slightly duplicates the schema parameter documentation, but the '요약은 호스트가 한다' sentence earns its place as it materially shapes how an agent should invoke the tool. No unnecessary bulk.

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?

For a single-optional-parameter tool with an output schema present and 100% schema coverage, the description is quite complete: it covers purpose, the non-calling-LLM behavior, the parameter, and gives an overview of the return (rules + required sections + document list + raw text in the prompt field). Since an output schema exists, the description need not detail the return structure further. Minor omission: it does not state prerequisites like the target folder needing to be scanned first, but the sibling scan_folder_structure makes this implicit.

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 coverage is 100%: the schema description for with_line_numbers already explains line-number attachment and the 'L14' citation rule, which is actually richer than the description's own '원문 줄 번호 부착 및 인용 규칙 추가 여부'. Both sources note the default of true. Per calibration, high schema coverage yields a baseline of 3, and the description adds only marginal value beyond what the schema provides.

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?

The description states a specific verb and resource: it returns a summary prompt that combines all target-folder documents with a standard report format. It also clarifies what the tool is not — it does not call an LLM and does not summarize itself, which sharply distinguishes it from every sibling (scan_folder_structure, extract_key_terms, validate_report_draft, etc.). An agent can tell this is the assembly step versus the read/extract/validate steps without opening the schema.

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 description clearly explains the tool's role in the workflow — it assembles instructions plus raw text for the host to consume, and explicitly states the LLM API is NOT called so the host does the summarizing. This tells the agent when to invoke it (as the prompt-building step before summarization) and why. It does not name specific sibling alternatives with when-not conditions, but the role is distinct enough among the siblings that this is a minor gap.

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