Package Pricing Lab

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

Annotations already establish readOnlyHint, idempotentHint, non-destructive and closed-world, so safety is covered. The description adds real value beyond that: results are deterministic arithmetic with the method explained, the variance buffer absorbs overruns from fixed-price adjustments, and delivery hours are used for internal cost/capacity only and 'never shown to clients'.

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 long sentences, but front-loaded: the first sentence states what is produced, the second what goes in, the third the return surface. The exhaustive output list is long yet justified because no output schema exists; only mild density cost.

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 9 parameters, nested objects and no output schema, the description correctly carries the return-value burden (tier contents, break-even and target price, rounded recommended price, margin, overrun headroom, capacity check). The one missing piece is any positioning against the sibling pricing tools.

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 89%, so the schema already carries the parameter detail, including defaults and the core/common/premium tier mapping. The description only restates this at a conceptual level ('fixed costs, owner pay, staff, delivery capacity') without adding syntax, units, or edge-case meaning, so the baseline of 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?

The description gives a concrete verb and artifact: turning a service business's work into three fixed-price tiers, and it enumerates the inputs (cost base, deliverables tagged core/common/premium) that define scope. It never names or contrasts with siblings like stress_test_package or compare_models, so an agent can't tell from the text alone which of these pricing tools to reach for.

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?

Applicability is implied by the detailed input narrative (requires a cost base plus leveled deliverables), but there is no explicit when-to-use or when-not-to-use statement, no prerequisites, and no routing to the alternative pricing/analysis tools. The closing note about adjustments being treated as included is a modeling behavior, not usage guidance.

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 cover safety (readOnly, idempotent, non-destructive, closed world), and the description adds real behavioral context beyond them: the current rate is always user-supplied and never assumed, the arithmetic is deterministic, the method is explained, and output is computed across several efficiency-gain scenarios with stated defaults. It stops short of describing failure modes or input validation behaviour.

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 the purpose, then inputs, then returns, so an agent can stop reading early. The middle and final sentences are dense lists, but each clause carries substantive information 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?

For a five-parameter nested tool with no output schema, the description supplies both the input contract and a detailed enumeration of return values (revenue/profit per job and month, earnings per hour worked, full-capacity ceilings, matching package price, findings), which is what the agent needs. The missing piece is routing guidance relative to its 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 only 40%, so the description must carry more weight. It does enumerate the main conceptual inputs (cost base, rate and unit, invoiced-hours share, one to three packages with price and effort, jobs per month) and states the efficiency-gain default of 0/15/30, but it says nothing about the optional inputs (staff cost, variable cost per hour, planned utilization, currency, direct-cost pass-through, hours_per_day), leaving a real gap.

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 (Compare) plus a precisely bounded resource: the economics of hourly/daily billing versus fixed packages for the same work. It is clearly separable from build_packages and stress_test_package, which construct and test a package rather than compare models.

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?

Implied usage is fairly legible (compare the economics of current time-based charging against packages), but there is no explicit when-to-use versus the siblings build_packages, package_scope_language and stress_test_package, and no when-not-to-use guidance. 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.

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

Annotations already establish readOnly/ idempotent/non-destructive, so the safety profile is covered. The description adds genuine content policy that annotations cannot convey: time-based or extra-charge wording is deliberately omitted and surfaced in notes, and voice defaults to 'we'. That is meaningful behavioral context beyond the structured fields.

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 output contents are front-loaded in a single dense sentence that enumerates each section, followed by one sentence on the omission policy and voice. Every clause maps to something the tool produces, though the run-on phrasing makes it harder to scan than a short bulleted structure would.

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 carries the burden of describing returns and does so thoroughly by naming every generated section and the paste-ready block per package. Combined with annotations covering safety, the only gap is guidance on insufficient input or the unmentioned 'ascending' flag.

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 67%, and the description's parameter-relevant content (voice 'we' default vs 'I' for solo, one-to-three packages) largely restates what the schema already documents. The 'ascending' parameter is not mentioned at all, so the description neither compensates for nor extends the schema beyond baseline.

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 gives a precise verb ('Write plain scope language') and resource (fixed packages), then enumerates the exact sections produced: included lines, adjustments line, out-of-scope lines, new-deliverable rule, worked examples, and a paste-ready block. It never references the siblings (build_packages, compare_models, stress_test_package), so an agent gets no explicit routing signal, which holds this at 4 rather than 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?

Usage is only implied: the tool generates scope language for one to three fixed packages, so an agent can infer it is used after packages exist. There is no statement of when to choose this over build_packages or stress_test_package, and no exclusions or prerequisites are given.

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 establish readOnlyHint/idempotent/non-destructive, so the safety profile is covered. The description adds genuine behavior beyond that: 'Deterministic arithmetic with the method explained,' plus the conditional nature of the month view (it only appears when sales_per_month is supplied) and the defaults applied for each scenario. It stops short of stating units or rounding of the returned figures.

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?

Purpose is front-loaded in the first sentence, followed by inputs and then the enumerated outputs. The output enumeration is long but justified because there is no output schema, so each clause earns its place; slightly denser than ideal but no filler sentences.

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 nested objects across 7 parameters, the description is the only source of return-value information — and it enumerates margin/profit per creep level, the zero-profit overrun, the target-margin price, and the full month view including break-even sales and reserve coverage. Nothing an agent needs to call or interpret the tool is 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 86%, so the parameters are largely self-documented and a baseline of 3 would hold. The description goes a step further by tying parameters to outcomes — sales_per_month enables the slow month view, cash_reserve drives the 'how many slow months the reserve covers' figure, and the creep/slow-drop defaults (10/25/50 and 40%) are surfaced in the narrative rather than left in 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?

Opens with a specific verb and resource: 'Show what scope creep and a slow month do to one fixed-price package.' It names the two stressors (creep, slow month) and the unit of analysis (one fixed-price package), which is enough to separate it from build_packages or compare_models without opening a 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 gives clear situational context — use it when you want to see the effect of scope creep and a slow month on a fixed-price package — and lists the inputs that enable each view ('optionally sales per month and a cash reserve'). However, it never names or excludes the sibling tools (build_packages, compare_models, package_scope_language), so the agent has no explicit routing rule among them.

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