InferCPI

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

With no annotations, the description carries the full burden, and it does disclose the pricing methodology (0.75 in + 0.25 out) and that capability scores appear only 'where matched' — real behavioral detail. It does not state that this is a read-only snapshot, whether results are cached or refreshed when, or what happens for models without an Epoch match beyond the vague qualifier.

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 lines, front-loaded with scope ('Every model priced in this week's index') followed by the returned fields, with no filler. The terseness is close to cryptic in places ('blended $/1M (0.75 in + 0.25 out)'), but every clause carries 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?

With no output schema and no annotations, the description does the necessary work of enumerating the returned fields and the price-blend formula. Given a zero-parameter read-only listing, the remaining gaps (freshness semantics, unmatched-model handling) are minor rather than blocking.

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, so the schema has nothing to document and the baseline is 4. The description's only parameter-adjacent content — the blend weights — describes output computation rather than inputs, which is fine given there are none.

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 makes the opaque name 'basket' concrete: it returns every model in this week's list-price index with provider, blended $/1M, and Epoch capability score. That is a specific resource with stated contents. It does not, however, distinguish itself from siblings like index_latest or model_price, which an agent must guess between.

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 is no statement of when to use this versus index_latest, index_catalogue, or model_price, and no exclusions or prerequisites. The agent must infer from the name and field list that this is a full-universe snapshot rather than a lookup or a series.

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?

With no annotations, the description carries the full behavioral burden. It usefully discloses methodology (fixed 8-task suite, OpenRouter default settings, charged cost, monthly cadence), but it does not explicitly state read-only behavior, freshness beyond 'monthly', or how medians are computed.

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 description is a single dense sentence that front-loads the core purpose ('Monthly measurement: dollars and tokens per completed task'). The parenthetical adds necessary methodology without becoming bloated.

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?

Given zero parameters, no annotations, and no output schema, the description does a reasonable job listing the returned metrics: dollars/tokens per task, reasoning share, accuracy, and medians. Some ambiguity remains around 'medians' and exact response fields, but it is sufficient for an agent to understand what the tool returns.

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 has zero parameters, so the baseline is 4. No parameter semantics are needed, and the description appropriately focuses on what the tool measures rather than inputs.

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 clearly states a specific measurement resource: monthly dollars and tokens per completed task, with supporting metrics like reasoning share and accuracy. It distinguishes itself from siblings such as model_price and speed by focusing on end-to-end cost-per-task under a fixed suite, though it does not explicitly name alternative tools.

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 gives no explicit guidance on when to use this tool versus alternatives such as model_price, speed, or open_vs_closed. Usage is only implied by the nature of the metrics, and there are no exclusions or routing instructions.

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?

With no annotations, the description carries the full behavioral burden. It usefully discloses ordering ('newest first'), event kinds, and returned fields, but does not state read-only safety, pagination behavior, rate limits, or authentication requirements.

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 description is a single dense sentence with front-loaded scope, followed by precise parameter semantics. Every clause (event kinds, returned fields, ordering, filters) earns its place with 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?

For a two-parameter read tool with no output schema, the description covers what is returned, how results are ordered, and how to filter. The remaining gap is explicit usage guidance versus sibling tools, which is minor but noticeable.

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 compensate. It explains 'since' as an ISO date and 'kind' as a filter to one kind, and the enumerated event types give concrete expected values for 'kind', making both parameters meaningfully clear.

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 a specific resource and behavior: 'Priced events in first-party price lists, newest first,' and enumerates the event kinds (launch, price_cut, price_increase, removed). It is clear what the tool returns and how results are ordered, though it does not explicitly contrast itself with siblings like model_price or index_series.

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 implied by the content (price change events), but the description offers no explicit when-to-use or when-not-to-use guidance relative to alternatives such as model_price or index_series. An agent can infer this is for change history, but routing is left largely to inference.

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?

No annotations exist, so the description carries the full burden. It does disclose the returned fields (signal/key, point count, first/last date, method string), which is valuable, but it never states that the call is read-only, whether the result is a full or filtered listing, or anything about result size or ordering.

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?

One short sentence, front-loaded with the resource scope and followed by the field list. Nothing is wasted and nothing requires scrolling.

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 parameters, no annotations and no output schema, the description must carry the return-value burden, and it does list the fields returned. It is nearly complete for this simple read tool, though it does not state that the listing is unfiltered or complete.

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, so there is nothing for the description to document beyond what the empty schema already shows. Baseline 4 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 is a noun phrase naming the resource ('every series the index publishes') and enumerates the returned fields, so an agent can infer this is a catalogue listing. However, it never states a verb (list/return/describe) and never distinguishes itself from siblings such as index_series or index_latest, which a reader must disambiguate on their own.

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 is no statement of when this tool should be chosen over index_series, index_latest, or basket. The reader must infer that it is a discovery step preceding those calls, but the description never says so or names any alternative.

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?

No annotations exist, so the description carries the full disclosure burden. It does well on payload content: it discloses that each row includes dates, the latest editorial note, and the licence/attribution line, which is real context an agent would not otherwise know. It says nothing about response size, cost, freshness cadence, or how 'latest' is determined, which leaves gaps for a tool with zero annotation coverage.

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?

A single dense sentence for the payload followed by a two-word directive ('Start here.') — front-loaded, no filler, and the actionable guidance sits at the end where it lands. The enumeration of series families is long but each item earns its place by telling the agent what it will receive.

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 must convey return content, and it does: series values across all listed families, dates, the latest editorial note, and the licence/attribution line. For a zero-parameter read tool that is close to sufficient. The remaining gap is response shape and volume — an agent cannot anticipate record count or format.

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, so the baseline is 4 and there is nothing for the description to disambiguate. It correctly implies an unparameterized fetch of all series at once, consistent with the empty 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?

Specific verb-less but precise resource statement: it returns the latest value of every index series, and it enumerates the distinct series families (list-price chained/composition/hedonic, paid price, Törnqvist, OpenRouter shares, TPI replica). That is far more informative than a title restatement. It does not explicitly contrast itself with the sibling 'index_series' (historical series) or 'index_catalogue', so sibling differentiation is only implied by the word 'latest'.

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?

'Start here' is an explicit entry-point recommendation, which is genuine usage guidance for an agent orienting itself in this tool set. However, it names no alternative and gives no when-not condition — nothing says to use 'index_series' instead when history or a single series is needed, so the alternative routing 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?

With no annotations, the description carries the full burden. It discloses the weekly granularity, return structure, and method string, which is valuable, but it does not state permissions, error behavior, or whether the operation is read-only.

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 description is compact and front-loaded with the return format, then defines each parameter efficiently. There is no wasted prose, and the signal examples are presented in a scannable way.

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 three-parameter retrieval tool with no output schema or annotations, the description covers the return shape, all parameters, and the key lookup path. It could be more complete only by explicitly distinguishing this from index_latest and indicating read-only behavior.

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 compensate, and it does: signal is illustrated with many values, key points to index_catalogue, and since is described as an optional ISO date cutoff. The signal list is not exhaustive due to the ellipsis, but the key parameter is well routed.

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 clearly states it returns the full weekly history of one series, including the return shape and method string. It distinguishes itself implicitly from index_latest by emphasizing 'full weekly history', but it does not explicitly name or contrast with that sibling.

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 useful context by pointing to index_catalogue for the key and explaining that since optionally cuts history. However, it does not explicitly state when to use this versus index_latest or other sibling tools, leaving the choice mostly implied.

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?

With no annotations, the description carries the burden. It usefully discloses that matching goes through the index's name normaliser and that multiple basket entries may be returned, but it says nothing about result ordering, ambiguity handling, or behavior when no match exists.

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?

A single front-loaded sentence that opens with what is returned and its units, then spends the remainder on input handling. Every clause carries information; 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?

For a one-parameter read lookup with no output schema, the description covers input forms, matching mechanics, and the multi-match return shape. Slightly more on ambiguity or missing-model behavior would close the loop.

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 compensate, and it does: the single `name` parameter accepts 'any common name' with three concrete format examples and explains that resolution is fuzzy via a name normaliser. Only the failure/ambiguity semantics remain unstated.

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 concrete resource and unit: blended list price per 1M tokens for a single model, resolved by name. It's clear this is a price lookup, though it doesn't name a sibling (e.g. cost_per_task or basket) to sharpen the boundary.

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 examples imply accepted input formats, which gently tells the agent how to call it, but there is no explicit when-to-use-this-instead-of guidance versus cost_per_task, basket, or index_latest. Usage is inferred rather than stated.

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?

No annotations are provided, so the description carries the full behavioral burden. It usefully discloses the formulas behind the comparison metrics, such as lab price divided by cheapest host and cheapest non-DeepSeek price divided by cheapest overall, but it does not describe the return shape, whether results are a table or scalar values, or any operational constraints.

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 content is relevant but packed into a single long sentence with multiple parenthetical metric definitions. It is not front-loaded with a high-level summary and could structure the list of computed metrics more clearly.

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 zero-parameter analytical tool with no output schema and no annotations, the description explains the metrics but omits the return format and when-to-use context. It is minimally adequate for understanding the calculations, but not complete enough for confident tool selection.

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 has zero parameters and an empty input schema, so there are no parameter semantics to explain. The baseline score of 4 applies because the description does not need to compensate for undocumented inputs.

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 comparative analysis: cheapest blended $/1M across closed vs open models, capability tiers, host discount, and DeepSeek floor lift. However, it is a dense noun phrase rather than a clear verb-led purpose statement, and it never distinguishes this tool from siblings such as model_price, cost_per_task, or index_latest.

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 describes the outputs the tool computes but gives no explicit guidance on when to use this instead of model_price, index_latest, cost_per_task, or other pricing/comparison siblings. There are no exclusions, prerequisites, or alternative-selection cues.

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?

With no annotations, the description carries the full burden. It does disclose the substance of what comes back — duty-cycle assumptions, which rental sources, break-even analysis, posted GPU-hour prices, and HF download counts — which is useful content transparency. However, it says nothing about behavior: whether these are static snapshots or live figures, data currency, or result format, all of which matter for a zero-annotation tool.

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 core value (cost per 1M tokens to leave the APIs) is front-loaded. The second sentence ('Plus posted GPU-hour prices...') reads as a tacked-on appendix and the first sentence is a dense run-on, which keeps it below a 5.

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 parameters and no output schema, the description must convey what the caller receives, and it enumerates the report's contents (per-source hardware costs, break-even, GPU-hour prices, download counts) in enough detail to set expectations. Missing freshness/format information is a modest gap given the tool's static-analysis nature.

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, so there is nothing for the description to disambiguate; baseline 4 applies. The description correctly does not waste space describing an empty 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?

The description makes clear this tool returns a self-hosting economics comparison: per-1M-token cost of local/rented/owned hardware versus API pricing of open and closed models. It names the resource (hardware vs API cost) precisely, but the bare name 'self_hosting' and the absence of any named sibling keep it short of a 5. It does not explicitly differentiate from model_price or open_vs_closed, which cover adjacent pricing territory.

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 implies the use case (deciding whether to leave APIs for hardware, needing break-even volume), but it never states when to call this instead of model_price, open_vs_closed, or cost_per_task. No exclusions, prerequisites, or routing guidance are given, so usage is inferred rather than instructed.

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?

With no annotations, the description carries the disclosure burden, and it does a solid job for a zero-parameter read: it defines the metric units, the routing configurations, the host attribution, and the fixed daily prompt that makes values comparable across days. It does not say how often the snapshot updates or whether the underlying data is cached, but for a no-arg read there is little actionable risk left undisclosed.

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?

One front-loaded sentence opens with the core scope ('Latest daily speed run'), then qualifies it with the route enumeration and metric definitions. The route parenthetical is dense but each clause supplies real information; 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?

No output schema exists, so the description must convey the return shape, and it does: one row per model and route with TTFT, throughput, and host. The only gap is the overall shape and size of the response (how many models or routes are covered, whether it is a list or a map), which an agent can only guess.

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, so the baseline is 4. The description adds no parameter semantics because none exist, and none are needed.

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 resource: the latest daily speed-benchmark run, with the measured metrics named explicitly (median time to first token, visible output tokens/sec, serving host) and the route breakdown enumerated. It is clearly distinguishable in content from siblings like cost_per_task or model_price, though it never names or contrasts those siblings, and it lacks an explicit verb such as 'returns'.

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 is no guidance on when to call this versus index_latest, index_series, or the other benchmark tools, and no mention of prerequisites or frequency of refresh beyond 'daily'. Usage is only inferable from the data being described, not stated.

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