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

Annotations declare readOnlyHint=true and destructiveHint=false, so safety is already covered. The description adds genuinely new behavioral context beyond that: visibility is role-scoped (agents see workspace totals plus their own row, humans see every agent), and in shadow mode limits are recorded as would_* values but not enforced — a crucial caveat an agent would otherwise misread as an actual block.

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?

Three short, front-loaded lines with no padding, and the shadow-mode caveat is placed where it will be read before calling. The opening line is a verbless noun list rather than a statement of what the tool does, which costs a little clarity for the brevity gained.

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 and read-only behavior is annotated, so the main remaining burdens are scope, caveats, and argument meaning — the first two are handled well. The definition is still incomplete for an agent that needs to decide between this and the budget/metrics siblings, and it leaves the window argument and cryptic terms like 'open circuits' and 'writes' unexplained.

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 must carry the meaning of the single 'window' parameter, and it never mentions it; the day/week/month aggregation scope and default are left entirely to the agent to infer from the enum. Worse, the text discusses 'mode="shadow"' as if it were a query argument, but no mode parameter exists in the schema, which can send the agent looking for an argument it cannot pass.

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 names the resource and its contents concretely — embedding tokens used vs quota, writes, would-be limits and open circuits — which is far more specific than the bare name 'get_usage'. It does not, however, differentiate itself from nearby siblings such as get_budget, get_metrics, or list_budgets, so the agent must infer which of these to call.

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 only implied: it tells the agent what rows it will see and what shadow mode means, which hints at when the tool is relevant (checking quota/limit status). There is no explicit when-to-use or when-not-to-use guidance and no mention of get_budget or get_metrics as alternatives for budget or metric questions.

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