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

Annotations already declare readOnly, idempotent, non-destructive, closed-world, and the description is consistent with them. It adds genuinely new context: deterministic computation, transient non-stored/non-logged input handling, a synthetic-inputs-only constraint, AP2 artifact export with execution_hash, and a verification receipts snapshot that is explicitly not a subscription.

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 heavily padded: it repeats the title, restates the compute-mode semantics verbatim from the schema, and embeds a page URL plus a full 64-character FV-status hash. The single most useful sentence (the purpose) is buried at the end instead of being front-loaded.

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 usefully enumerates the return fields (consistent, grand_total_delta, subtotal_deltas, tax_total_delta, parse_error, etc.) and covers provenance chaining and privacy. The one remaining gap is the expected shape of the input document, which is deferred to the upstream art-293 validator.

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 all four parameters are already documented in the schema, including the compute enum and the parent_hashes/parent_tool_ids pairing. The description's compute-mode paragraph largely restates the schema and adds no syntax or format detail for document/rounding beyond 'supplied rounding convention'. Baseline 3 is appropriate.

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 'Summary' sentence states a specific verb and resource: recompute per-line VAT, category subtotals, tax total and grand total from line items under a rounding convention, then compare against document-asserted amounts. That clearly separates it from siblings like validate_einvoice_format or route_einvoice_jurisdiction_mandate. The problem is placement — the actual purpose arrives only after a wall of compute-node boilerplate.

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?

Implicit sequencing is conveyed by 'Consumes upstream artifacts from: art-293-einvoice-format-validator' and 'Output feeds: art-295-einvoice-jurisdiction-mandate-router', which tells the agent where this sits in the chain. But there is no explicit when-to-use/when-not statement or comparison against alternatives such as validate_einvoice_batch or recompute_* siblings.

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