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

Goes well past the annotations: it discloses that the ID is retained, saved evidence is reused, certain outputs are immutable, and the operation is asynchronous and must be polled. The 'Reuse the key on retry' line adds idempotency-key retry semantics; it sits in mild tension with idempotentHint=false but is not a contradiction, since a fresh key still produces a new edit.

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?

Six short sentences, action and in-place scope front-loaded, then preconditions, exclusions, polling, and retry. Nothing is padding; each sentence carries a distinct operational fact.

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?

The tool is a destructive in-place mutation with an output schema, and the description covers preconditions, immutability cases, async polling, and retry behavior. It stops short of saying whether the prior report content is overwritten or versioned, and says nothing about required permissions.

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 only 20% (only organization_id is documented), so the description has to carry the load. It refers to all four required params in prose — the prompt, the existing output (output_id), the session via inspect_research_session, and the idempotency key — but adds no format or semantics beyond their names.

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 resource: 'Apply the requested prompt to an existing report or PowerPoint in place.' The words 'existing' and 'in place' distinguish it from generation/regeneration siblings such as generate_research_output, though no sibling is named explicitly.

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?

Gives real preconditions ('Read the output first'), an explicit exclusion ('default reports and Highlights cannot be edited'), and a follow-up path ('Poll inspect_research_session view=job and read the original output after success'). It never states the edit-vs-generate decision rule against generate_research_output, which keeps it below a 5.

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