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

Annotations already declare mutation, idempotency and non-destructiveness, and the description adds genuinely new facts: a per-call cost of $0.001 with no charge on failure, and that assigning a role steals it from whichever account previously held it. That side effect on a different record is the most important non-obvious behavior and is disclosed; what remains unstated is return shape, batch atomicity, and what happens if a role is requested while the type forbids it.

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?

Front-loaded with the operation, then the mutable fields, then the role explanation — good ordering with no filler sentences. However, the 20-value role list duplicates the enum already present in the input schema, adding length without adding meaning.

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 nested-array upsert with no output schema and no annotations covering field semantics, the description supplies the essential preconditions and side effects. It is slightly incomplete on write-failure behavior beyond billing, on whether batches are atomic, and on what the call returns for each account.

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 0%, so the description carries the load, and it does so for the hard parts: it explains that a role is a posting account, enumerates the role semantics, and gives update preconditions for type and archived. The trivial fields (name, bic, new_code) are only named without added meaning and the 100-item batch cap is never mentioned, so it stops short of full compensation.

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 pair and resource ("Add accounts or change them") plus the key matching semantics ("matched by code"), which immediately distinguishes it from the sibling list/read tools such as books_list_accounts. An agent knows this is an upsert on the chart of accounts 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?

Usage context is implied by the operation itself and by conditional constraints ("type (only while unused)", "archived (only with no balance and no role)"), but there is no explicit when-to-use guidance, no mention of prerequisites like looking up existing codes via books_list_accounts, and no stated alternative for other account mutations.

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