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

Annotations provide idempotency and non-destructiveness, so the description doesn't need to restate those. The description adds valuable behavioral details beyond annotations: it writes 'user overrides that AI enrichment never overwrites,' re-derives 'each bookkeeping journal entry,' and returns 'how many applied plus per-transaction failures.' These are useful side effects and return semantics not present in the annotations. A bit more could be said about permission requirements, but the given coverage is strong.

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?

The description is three sentences with no filler. The core action and scope are first, and behavioral/return guidance and alternative guidance follow; every sentence adds value. It is dense and front-loaded while still maintaining clarity.

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 batch mutation tool with 3 params, an output schema, and annotations, the description covers the necessary operational details: batch size, account path format, behavioral outcomes, return/errors, and the replacement pattern. It doesn't explicitly mention when not to use it versus the other batch sibling, but it is still complete enough for an agent to invoke the tool correctly for the described scenario.

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 33%, so the description carries extra weight. It does add meaningful semantics: the example 'expenses/meals-and-entertainment' explains glAccountPath's expected format, and 'up to 500 transactions' re-frames transactionIds max. But it doesn't help much for companyId or the wrapper structures (already in schema), so it partially mitigates the 33% coverage gap. The balance makes it a 4.

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 states a specific verb ('Book'), the resource (transactions to a chart-of-accounts account), a clear scope (up to 500 transactions in one call), and how the target account is expressed ('full account path'). It also mentions the return type (applied count + failures). It distinguishes from the single-transaction sibling by explicitly saying 'Use this instead of calling transaction_categorize_transaction_by_account in a loop.' Agent can immediately tell what this tool does and which sibling it replaces.

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 gives an explicit alternative guidance: use this instead of looping over transaction_categorize_transaction_by_account, which is very helpful. However, it does not address the sibling transaction_batch_categorize_transactions, which could be confused with this batch-by-account tool. It is clear about the 'same account' condition but could say when the other batch tool is needed. Because the primary alternative is named and the when/when-not is explicit, this is a 4.

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