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

Beyond the readOnlyHint/openWorldHint annotations, it discloses the model's provenance (same math as Hemrock's Ecommerce Forecasting Tool spreadsheet, checked against the workbook), the defaulting behavior, and that the result echoes every input used so defaults can be surfaced to the user. It also explains the quota surface — free, no calculations consumed when signed in, a few free runs a day without an account — and that webUrl opens the model live for editing and Excel download. That is the kind of context annotations cannot carry.

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 trigger and mechanism, then mode selection, then return shape, then billing — a sensible order. It is dense rather than padded, and the parenthetical curve explanation is genuinely clarifying. It does restate some parameter detail already present in the schema (manualRetentionCurve semantics, months range), which is the only thing keeping it from 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?

This is a one-parameter tool hiding 24 nested inputs and no output schema, so the description has to carry the return shape — and it does: each line by month, the retention curve used across all 120 months, lifetime orders per customer, LTV, and a webUrl. Combined with the defaulting rule and the statement that only user-supplied inputs need sending, an agent has everything required to call it correctly and report results.

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 would be 3, but the description adds real meaning on top of the schema: it explains the indexing convention for manualRetentionCurve (month 1 is the first order, 0.3 in month 4 means 30 of 100 reorder), how missing months are treated, and when to prefer manual entry over the calculated curve. It also clarifies the months range/default and that ltvMonths caps lifetime orders and LTV independently of forecast length. Mostly duplicative of the nested schema descriptions, but not purely duplicative.

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 — build an ecommerce/DTC forecast on customer cohorts — and immediately explains the mechanism (each month's new customers form a cohort that reorders along a retention curve) plus the outputs (orders, revenue, cost of sales, marketing, operating income, contribution margin per order, traffic). That cohort framing clearly separates it from saas_forecast_compute, unit_economics_compute and fund_economics_compute without needing to name them.

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 opening 'Use when the user wants an ecommerce or DTC forecast built on customer cohorts' sets a clear trigger, and it goes further with a mode-selection rule: 'Use the entered curve whenever the user describes retention month by month or has cohort data.' It also tells the agent to send only user-supplied inputs and to relay webUrl to the user. The one gap is that it never contrasts itself with the sibling forecast tools, so an agent choosing between ecommerce, SaaS and unit-economics forecasts gets no explicit exclusion rule.

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