Stackup: Fannie, Freddie and HUD multifamily loan sizing

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

Annotations already cover safety (readOnly, idempotent, non-destructive, closed-world), so the description's job is to add context beyond that — and it does, enumerating the return payload (fit/no-fit with reason, max loan by LTV/LTC and DSCR, binding constraint, payment incl. HUD MIP, term, warnings, ranked by max loan). This is valuable since there is no output schema.

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?

Two sentences, no filler, and the program scope is front-loaded before the payoff list. The first sentence is long and clause-heavy (nested program enumerations), which slightly hurts scannability, but every element earns its place.

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 23-parameter computation tool with no output schema, the description is complete: it names the programs, the eligibility verdicts, the computed outputs, the ranking, and the excluded product types. An agent knows what it will get back and what the tool will not tell it.

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%, including defaults, ranges, enum meanings and conditional usage ('used only when noi is omitted'), so the schema already carries parameter meaning. The description adds no parameter-level guidance of its own beyond mentioning HUD MIP, so baseline 3 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?

Specific verb ('Runs the multifamily.online comparison app'), precise resource (one 5+ unit building), and explicit program scope (Fannie DUS/Small, Freddie Optigo Conventional/Small, HUD 223(f)/221(d)(4)/223(a)(7)). It also states which loan types are out of scope, so an agent can tell it apart from get_program_terms and get_treasury_yields 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?

The description makes the use case clear (size/compare one deal across agency and HUD programs) and rules out SBA, bank, CMBS, bridge and DSCR products. It does not, however, explicitly route the agent to get_program_terms or get_treasury_yields, so the sibling relationship is left implicit.

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 already declare readOnlyHint, idempotentHint, non-destructive, and closed-world, so safety is covered. The description adds meaningful context beyond that: it discloses the data provenance ('as multifamily.online lists them', 'the agency source document with its date and URL'), which tells the agent the values are third-party-sourced and dated rather than computed live. It does not discuss caching or how often the source data is refreshed.

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?

Two sentences: the first front-loads what is returned, the second enumerates covered programs. The program list is long but functional since it tells the agent the tool's full coverage. Minor redundancy between 'all of them' and the enumerated list, but no filler.

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 compensates by enumerating the returned fields, and the read-only annotations cover the safety profile. The main residual gap is the lack of routing guidance relative to compare_programs, which matters for a tool whose sibling also spans multiple programs.

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% and the single enum parameter is fully documented with its default, so the schema carries the semantic load. The description's 'one ... or all of them' merely restates the schema's default behavior and adds no new parameter meaning.

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 a specific resource (published loan-program limits) and enumerates the exact fields returned (max LTV/LTC, min DSCR, term, amortization, prepayment, recourse, HUD insurance, source document). An agent immediately understands this is a reference-lookup tool. It stops short of distinguishing itself from the sibling compare_programs, which an agent could easily confuse with a multi-program query.

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?

The phrase 'for one multifamily loan program or all of them' implies the two usage modes, and the coverage list signals which loan types are in scope. However, there is no explicit when-to-use versus compare_programs or get_treasury_yields, and no stated exclusions or prerequisites.

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 already declare readOnly, idempotent, non-destructive, closed-world behavior, so the safety profile is covered. The description goes beyond them by disclosing that each value carries an as-of date, that week- and month-earlier readings appear only 'when available', and that a freshness status and publisher link are included — real signals about possibly stale data and variable presence of fields. It stops short of stating units (e.g. percent) or whether multiple entries per tenor are returned.

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?

It is a single front-loaded sentence that leads with the instrument tenors and then the metadata fields, with no filler. It is dense — the 'each with its...' clause runs long — but every element listed corresponds to a returned field, so nothing is wasted.

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?

There is no output schema, so the description carries the burden of describing the shape of the result, and it does so by enumerating series, as-of dates, prior readings, freshness and source links. Given zero parameters and full annotation coverage, this is nearly sufficient; only units/format of the yields and the precise structure of the response remain unstated.

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?

The tool takes zero parameters, which is the baseline-4 case; there is nothing for the description to disambiguate. The description correctly implies a no-argument 'fetch latest snapshot' call by never referencing any filter or option.

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 a specific resource (5-, 7- and 10-year Treasury par yields plus overnight SOFR) and enumerates the accompanying fields, so an agent knows exactly what this returns. It also implicitly separates itself from the program-focused siblings by being macro-rate data, though it never names compare_programs or get_program_terms to state the boundary explicitly.

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 is only implied: the phrase 'that the multifamily.online app builds its example rate on' hints this is the benchmark-rate lookup to use when explaining pricing context. There is no explicit when-to-use statement, no exclusions, and no mention of the sibling tools, so the agent must infer the routing decision.

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