Start an analysis

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

Annotations already declare readOnlyHint=true and destructiveHint=false, so safety is covered; the description usefully adds the ordering constraint that it must be called first and that it returns steps. It does not mention the consequence of calling it repeatedly (idempotentHint=false implies each call starts a fresh analysis, potentially discarding prior workflow state), which is the one behavioral trait an agent most needs here.

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?

It is front-loaded with the imperative call rule, then mode discrimination, then return value, with no filler. The only waste is the partial duplication of the mode semantics that already appear in the schema, which makes the middle sentence heavy.

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?

With no output schema, the description correctly compensates by noting it 'returns the steps to follow', and it covers the required question/mode decisions fully. What is missing is the lifecycle context: what happens to an in-progress analysis if the tool is called again, and whether the returned steps must be passed to run_test, describe, or another sibling.

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%, so both required parameters are documented in the schema itself. The description's mode definitions ('caused by something', 'test a claim', 'describe, summarise, visualise, estimate or plan') add a little interpretive help beyond the enum descriptions, but the two texts largely overlap, so baseline 3 applies.

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 frames this as the mandatory entry point for every new question and states it returns the steps to follow, which distinguishes it from executable siblings like run_test or describe. However, it never states plainly what the tool does with the question (it classifies and routes into an analysis workflow), so the underlying action must be inferred from the mode rules.

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?

It gives a clear invocation rule ('call this first for every new question') plus explicit decision criteria for choosing between the two modes, and tells the agent to ask the user when ambiguous. It does not reference any sibling tools such as skip_step or workflow_status, so the agent gets no guidance on how this tool relates to the rest of the workflow after the initial call.

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