export_research_package

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

Annotations already declare readOnlyHint and idempotentHint, so the safety profile is covered. The description adds real context beyond that: the output is checksummed, and the tool serializes supplied data rather than deriving it ('serializes supplied evidence and metadata rather than launching predictors or docking') – an important expectation that a missing section cannot be regenerated. It says nothing about failure modes when required sections are absent, which keeps it below 5.

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?

Three short sentences, front-loaded with what the tool produces, then when to use it, then a scope boundary. No filler and nothing repeated from structured fields.

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 7-parameter tool with deep nested objects, the description covers purpose, timing, the sibling alternative, and the non-computing nature; an output schema exists so return values need no explanation, and the schema documents every field. The remaining gap is behavior when the supplied snapshot is missing a section flagged by an include* boolean, which cannot be inferred from either structured field.

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%, including per-flag text ('Whether to include supplied ... artifacts'), so the schema carries the parameter burden and baseline 3 applies. The phrase 'selected sections' hints that the include* booleans gate package contents, but it adds no format, default, or interaction detail beyond what the schema already documents.

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 artifact ('Build a checksummed research ZIP') plus the inputs it consumes ('supplied package snapshot and selected sections'). The closing clause explicitly separates it from compute tools ('rather than launching predictors or docking') and from generate_report, so an agent can place it 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?

'Use after reports and artifacts are ready for handoff' gives a clear precondition, and 'generate_report exports individual report files' names the nearest alternative with the distinction that selects between them. It does not cover the other export siblings (export_candidates, export_workflow_trace), so the when-not guidance is incomplete rather than absent.

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