analysis_plan

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

With no annotations provided, the description carries the full burden and discharges it richly. It discloses deterministic rule-based behavior ('零 LLM、零模糊匹配'), fallback plan when no intent matches, multi-intent priority by table order, specific error codes (E1001) for invalid inputs, and the behavior of ignoring invalid column_hints references while noting them in the summary.

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 description is long but well-structured with clear sections for parameters, intent table, return value, and examples. Most content earns its place; minor meta-information like 'v1.2.0 方案 B 落地' and '与 statlab_mcp/docs/design/11_analysis_plan.md 同步维护' is useful context but not strictly needed for invoking the tool.

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?

Given the tool's complexity, the absence of an output schema, and zero schema descriptions, the description is remarkably complete. It covers all parameters, the full intent list, fallback behavior, the result structure with keys like chosen_methods and tool_calls_plan, a summary template, and two representative examples. An agent has enough to select and call this tool correctly without external documentation.

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 0%, so the description must compensate fully. It explains each parameter's meaning, requiredness, mutual exclusivity constraints, valid value domains for column_hints ('数值|类别|日期'), and error behavior for invalid values. This goes well beyond the raw type-only schema.

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 opens by identifying the tool as an orchestration-layer analysis plan generator ('编排层 · 分析计划生成') and immediately clarifies it converts a decision tree into a deterministic rule tool that '只出计划不执行' (only plans, does not execute). This clearly distinguishes it from all the sibling execution tools like hypothesis_test, linear_regression, and describe_statistics, which actually perform the analyses.

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 gives strong usage boundaries: it is for generating plans, execution is left to the outer agent calling first-layer tools, and it documents parameter preconditions (question required, file_path/inline_data mutually exclusive, both missing is legal, both provided triggers E1001). It stops short of naming specific sibling alternatives for when-to-use-not, but the orchestration-versus-execution contrast is explicit.

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