prontofax

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

Annotations already cover the safety profile (readOnly, idempotent, closed-world, non-destructive), and the description still adds real behavior: results carry the official source and last-verified date, and the tool fails closed by returning nothing rather than guessing. Missing only operational notes such as latency or whether a fuzzy agency string can return multiple candidates.

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 purpose and the city/state constraint are front-loaded, but the long roll-call of covered states and agencies consumes substantial text for scope that could be summarized more compactly. Every clause is informative, yet the enumeration is token-heavy relative to the decision it supports.

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 compensates by describing what comes back (number, source, verification date) and the empty-result behavior, which is what an agent needs before calling. Given the closed-world read tool, only a note on ambiguous agency matches is arguably missing.

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 100%, so the baseline is 3, and the description earns above that by explaining that agency is matched from the user's own words and that city/state are the disambiguators required for SSA and Medicaid. It adds the rule that state is the only field read as a bare code, which the schema only hints at.

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?

"Find the fax number for a US state or federal agency" is a specific verb+resource, and the enumerated coverage (SSA field offices, MassHealth, state UI/Medicaid agencies) makes the scope unmistakable. It never names the sibling send_fax, so the lookup-versus-send distinction is left to be inferred from the verb.

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 states the applicability conditions clearly, including the negative boundary "It has no numbers for any other agency" and the rule that SSA and Medicaid lookups require a city or state. It does not tell the agent when to route to send_fax instead, or what to do after a miss beyond the fact that nothing is returned.

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

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

Annotations only give the generic non-read-only/open-world profile; the description adds the material behavior: nothing is sent or charged until the user confirms and pays in the form, and billing happens only on confirmed delivery. It also discloses the host-dependent fallback (returns a link) and the constraint that chat attachments cannot be passed in — exactly the surprises an agent would otherwise hit.

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 purpose and mechanism, then cost, then the attachment caveat and fallback — a logical order with no filler. It runs slightly long at six sentences, and the two attachment sentences could be merged, but nothing is redundant.

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 form-driven, human-in-the-loop tool with no output schema, the description covers everything needed to set expectations: where the document goes, what the user pays, when charging occurs, and what happens on hosts without form support.

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 each of agency, language, and fax_number is already documented, including that the number may come from an attached letter. The description adds no parameter-level meaning beyond the schema's own note, so the baseline applies.

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 opens with a concrete verb and resource — "Start a fax to a US fax number" — and immediately clarifies the interaction model (a form opens in the chat). That is enough to distinguish it from the only sibling, lookup_fax_number, without reading either 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 explicit: no account needed, US numbers only, $2.99 flat up to 25 pages, and a browser-link fallback for hosts that cannot render the form. The one gap is that it never tells the agent to call lookup_fax_number first when the destination number is unknown, which is the obvious companion action.

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