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

Annotations are all negative signals (readOnlyHint=false, idempotentHint=false, destructiveHint=false), so the description carries the full behavioral burden. It discloses two non-obvious traits: revisions require the exact draftId plus expectedRevision pairing, and stale writes are rejected — a concurrency guarantee that materially changes how an agent must invoke an update. This is exactly the kind of context annotations cannot express.

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 sentences with zero filler: scope, persisted state, and concurrency rule each earn their place, and the primary purpose is front-loaded in the first clause. The ~48-word length is appropriate because the schema already carries parameter-level detail, so the description does not repeat it.

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 mutating tool with zero required parameters and no output schema, the definition thoroughly covers the invocation contract: what state is saved, how revisions behave, and when it does not apply. The one gap is the return contract — it never states that a create returns a draftId, which a downstream workflow would need for later revisions or a subsequent create_study call, so the agent must infer it.

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% with rich per-parameter descriptions (e.g., groupIds is a 'Legacy alias for audienceIds', methodId defers to list_research_methods), establishing a baseline of 3. The description adds framing on top: it groups the 13 parameters into a coherent persisted state bundle and reinforces the draftId/expectedRevision dependency that underlies the stale-write guard.

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?

Opens with a specific verb-resource pair — "Creates or checkpoints an unfinished Quick or Custom Study draft" — and delimits scope with "without starting research," which distinguishes it from the create_study sibling. The sentence enumerating exactly what gets saved (objective, Audiences, method, questions, sources, planner step) further pins down what the tool owns.

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 establishes clear context: this tool persists incomplete planner state for later restoration, and "without starting research" signals a when-not condition. It stops short of naming the explicit alternatives (create_study for launching a study, list_study_drafts for reading drafts), so routing to siblings relies on inference rather than an explicit when-to-use/when-not-to-use statement.

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