core.explain

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

Annotations already declare readOnly/idempotent/closed-world, and the description still adds meaningful behavior: numbers are passed through verbatim, the prose only narrates the trace, and long values are truncated at 300 characters with the original preserved in `result`. It omits failure behavior (invalid tool name, non-read-only tool rejected) and any cost/latency expectations.

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 the mode and the operation, then output shape, then fidelity guarantee, then truncation. Slight redundancy between the fidelity sentence and the truncation sentence, but nothing is padding.

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?

An output schema exists, yet the description still usefully enumerates the returned fields, and it discloses the read-only restriction and truncation rule an agent needs to interpret results. Missing only edge-case handling (invalid/nested tool, error surfacing) for a meta-tool with a free-form nested `arguments` object.

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 coverage is 0%, so the description must carry the load and largely does: `tool` is the read-only tool to run, `arguments` is that tool's argument object, and `lang` selects ko or en for the prose. It does not clarify the tool-name format (dotted identifiers like accounting.balance) or whether `arguments` must match the target tool's 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?

States a specific action: execute a read-only tool with `arguments` and return a plain-text narration of the formula, inputs, steps, result, applied policy, and statutory basis. The 'explanation mode' framing makes it distinguishable from plain calculation, but it never names the obvious sibling (core.calc) as the non-explaining alternative.

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 does constrain usage to read-only tools, which is a real when-not-to-use signal for mutating siblings. However there is no explicit guidance on when an agent should prefer core.explain over core.calc or over calling the target tool directly, so the routing decision is left to inference.

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