Diff workbooks

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

Annotations already declare readOnlyHint=true and openWorldHint=true, so the safety profile is covered; the description adds real context beyond that — files are uploaded to and parsed on the Basis API, capped at 20 MB each, consume the API key's parse quota, and are not stored. These quota and data-handling notes are exactly the kind of behavioral detail annotations cannot express, so the description earns credit above the annotated baseline, though it says nothing about failure modes (e.g. mismatched sheet sets or unsupported file types).

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?

Four sentences, front-loaded with purpose and return shape, then usage routing, then invocation and cost constraints — no filler. The dense return-shape enumeration earns its space because no output schema exists.

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 supplies the JSON keys and the change-semantics needed to interpret results, and it covers cost, size limits, and source exclusivity for a 7-parameter tool. The only gap is that it does not state what happens when the two workbooks differ structurally (differing sheet sets) or how large diffs are returned.

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 the meaning of each of the seven parameters is already documented. The description's 'provide exactly one of before_path/before_url/before_base64' rule adds genuine constraint the flat schema does not express (no oneOf and required is empty), which is a modest but real value-add over the 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 with a specific verb and resource ('Compare two .xlsx files cell by cell') and immediately defines the output shape {summary, added, removed, changed, format_changed}. The matching rule (sheet title, cell ref) and the value/formula vs format-only split further pin down what the tool actually computes, so an agent can distinguish it from audit_workbook or recalc_workbook without opening a 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?

It gives a concrete usage scenario ('before and after an edit') and routes the agent away from two named siblings with explicit conditions: audit_workbook for problems inside one workbook, restore_session for undoing session edits. The exactly-one-of source constraint adds an additional structural rule for invocation.

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