Read-only flight fares for AI assistants: the cheapest fares a2bflight has seen on thousands of routes, one-way and round trip, each with the day it was seen. Three tools, no auth, no key.
Scored across 3 tools
Each tool has a clearly distinct role: find_airports resolves codes, cheapest_routes browses routes by origin/destination, and route_fares inspects a single airport pair. The descriptions make the broad-vs-specific boundary explicit, so misselection is unlikely.
All names use consistent lower snake_case, which is predictable and readable. The grammatical pattern is not strictly verb_noun throughout (find_airports vs. cheapest_routes and route_fares), but the deviation is minor.
Three tools are well-scoped for a read-only cached fare browser: resolve airports, discover cheapest routes, and inspect a specific route. Each earns its place without redundancy.
The surface covers the full read-only fare lookup lifecycle: code resolution, route discovery across origins/destinations, and detailed cached fares for a specific airport pair. route_fares even handles empty caches by triggering pricing, so there are no obvious dead ends.
The cheapest cached fare on every route we hold from an origin (or into a destination), cheapest one-way first, each with its cheapest round trip of a 3 to 21 day stay in return when we hold one. Filter by a maximum price and a departure month here rather than by reading the list; both apply to the one-way. Prices are one adult, economy, as last seen on the date in checked — observations, not quotes.
| Name | Required | Description | Default |
|---|---|---|---|
| to | No | destination IATA airport or city code | |
| from | No | origin IATA airport or city code (NYC covers JFK, LGA and EWR); give from, to or both | |
| limit | No | how many routes to return, 1 to 20; 10 when omitted | |
| month | No | YYYY-MM: the cheapest fare departing in that month; needs from | |
| max_price | No | keep only fares at or under this amount, in the currency the results carry |
| Name | Required | Description |
|---|---|---|
| note | No | |
| routes | Yes | |
| currency | No |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
Annotations already declare readOnlyHint, openWorldHint=false and destructiveHint=false, so safety is covered. The description adds genuinely non-obvious behavior: prices are cached observations 'as last seen on the date in `checked`', not live quotes, and the round-trip component only appears 'when we hold one' — both material to how an agent should present results.
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 what is returned before the filtering advice and the pricing caveat. The opening sentence is long and parenthetical-heavy, but every clause carries information and nothing is redundant with the schema.
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 an output schema present, the description is not obligated to detail return values, yet it still characterizes the shape of a result (one-way plus optional round trip). Combined with the staleness caveat and filter semantics, an agent has enough to call and interpret the tool; only sibling routing is left implicit.
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, but the description goes further: it explains that `from` and `to` are interchangeable anchors ('give from, to or both'), that the month/max_price filters apply to the one-way leg only, and that the `return` companion is a 3-to-21-day stay. That clarifies cross-parameter interaction the schema does not express.
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 verb and resource: the cheapest cached fare on every route from an origin (or into a destination), with one-way first and a round-trip companion in `return`. The scope is unambiguous and clearly a list-over-routes tool rather than a single-route lookup, though it never names `route_fares` or `find_airports` to sharpen the contrast.
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 gives one useful directive: filter by max price and departure month via the parameters 'rather than by reading the list.' That is guidance about how to invoke, not about when to choose this tool over its siblings, and no prerequisite or exclusion conditions are stated.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Look up airports and city codes by name, or list the airports near a code or a lat,lon point. Use it to turn a city name into the IATA code the other tools take.
| Name | Required | Description | Default |
|---|---|---|---|
| near | No | an IATA code, or lat,lon, to list the airports near it; give this or query | |
| query | No | a city or airport name or the start of one; give this or near |
| Name | Required | Description |
|---|---|---|
| places | Yes |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
Annotations already declare readOnlyHint=true, destructiveHint=false, and openWorldHint=false, so the agent knows this is a safe, closed-domain read. The description adds the dual-mode behavior (name lookup vs near-code/coordinates), which is useful, but says nothing about result volume, ranking, or limits on the proximity list.
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 tight sentences, zero filler. The lookup-vs-list distinction is front-loaded and the downstream-purpose sentence is the key routing information.
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, so return-value documentation is unnecessary. Both input modes and the practical purpose are covered, which is everything an agent needs to select and call this tool against its two-parameter, zero-required schema.
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 the schema already documents both 'query' and 'near' including their either/or relationship. The description adds only the framing of turning a city name into an IATA code, which clarifies intent but no syntax or format detail beyond the 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 specific verbs (look up, list) over a specific resource (airports and city codes) and covers both operating modes: name lookup and proximity listing. It also positions the tool relative to its siblings by naming the downstream purpose (producing the IATA code other tools take).
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?
Gives a clear context of use: resolving a city name into the IATA code that the sibling tools consume, plus the two query modes. It stops short of explicitly naming cheapest_routes or route_fares as the consumers or stating when to prefer one mode, so it's clear context without explicit alternatives.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Everything we hold for one airport pair: the cheapest one-way, the cheapest round trip of a 3 to 21 day stay, the cheapest one-way by departure date and by month, and the cheapest round trips. If we hold nothing yet, it starts pricing the route and says so; ask again a minute later.
| Name | Required | Description | Default |
|---|---|---|---|
| to | Yes | destination IATA airport code | |
| from | Yes | origin IATA airport code (an airport, not a city code) | |
| month | No | YYYY-MM: keep only departure dates in that month |
| Name | Required | Description |
|---|---|---|
| to | Yes | |
| from | Yes | |
| dates | No | the cheapest one-way fares by departure date, cheapest first |
| months | No | the cheapest one-way fare in each month |
| status | Yes | held: fares below; pricing: we just started pricing this route, ask again in a minute; none: we hold no fare for it |
| returns | No | the cheapest round trips held, any stay, cheapest first |
| cheapest | No | the cheapest one-way |
| currency | No | |
| page_url | Yes | |
| cheapest_return | No | the cheapest round trip of a 3 to 21 day stay; absent when we hold none, and page_url then offers a live return search |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
Annotations already cover read-only, open-world and non-destructive safety. The description adds genuinely valuable behavior: an empty-cache case that triggers background pricing and returns only a 'pricing started' notice, with a retry-after-a-minute hint. That async/latency trait is the kind of thing an agent cannot get from the schema or annotations.
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, front-loaded with the return contents, and the retry caveat comes last. The middle enumeration ('cheapest one-way ... and the cheapest round trips') is somewhat list-heavy but each clause is informative rather than 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?
An output schema exists, so return values needn't be spelled out, and the definition still covers the important edge case (no cached data yet). Combined with the safety annotations, an agent has enough to call it correctly. The only gap is the lack of routing guidance among siblings.
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 from/to and month are already documented in the schema. The description adds a little framing ('by departure date and by month') but no syntax or format meaning beyond the schema's 'YYYY-MM' note. 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 description states concretely what the tool returns for one airport pair: cheapest one-way, cheapest 3-21 day round trip, cheapest one-way by departure date and by month. That is a specific, enumerable resource. It does not name or contrast against the siblings cheapest_routes or find_airports, so it stops short of 5.
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?
There's useful situational guidance: if nothing is cached the tool starts pricing and the caller should ask again in a minute. But there is no explicit when-to-use-vs-alternatives routing (e.g. when to prefer cheapest_routes), so the agent must infer the boundary from the siblings' names.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Recent tool additions, removals, and schema changes observed during successful MCP inspections.
cheapest_routesfind_airportsroute_fares
Add one secure layer between your agents and this server.