dml_sensitivity

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

Annotations declare readOnlyHint=true, so the safety profile is covered. The description adds substantial behavioral context beyond that: assumptions (unconfoundedness, overlap, Neyman-orthogonality, nuisance convergence), failure modes (overlap failure inflating variance/bias), and explicit pre-conditions. This goes far beyond the minimal read-only hint, providing actionable behavior expectations.

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 organized with labeled sections (Assumptions, Pre-conditions, Failure modes, Alternatives, Typical minimum N), which makes it scannable and front-loaded with the core purpose. It is relatively long, but each section adds essential information and there is minimal redundancy. A minor deduction for length and the somewhat cryptic 'Validation: certified parity evidence' sentence, which adds little clarity.

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 the complexity of a DML sensitivity tool, the description is remarkably complete. It covers assumptions, pre-conditions, failure modes, alternatives, and a sample size guideline. Since an output schema exists, return values are covered. There is nothing an agent needs to know to invoke this tool correctly that 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 description coverage is 100%, so every parameter is documented in the input schema. The main description does not add any parameter-level semantics beyond what the schema already provides; it mentions q and cf in the context of the sensitivity logic, but that information is already present in the schema. The description thus adds no extra value for parameters, so the 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?

The description states a specific verb ('Compute') and resource ('DML-OVB sensitivity') targeted at a 'fitted DML CausalResult', which is precise and distinguishes it from the many sibling sensitivity tools (e.g., sensitivity, sensitivity_dashboard, sensemakr). It clearly conveys the tool's core function without ambiguity.

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 provides explicit pre-conditions (covariates, treatment, outcome, instrument for IV variants), failure modes (overlap failure near 0/1 propensity) with remedies, and names alternatives (sp.tmle, sp.auto_cate, sp.causal_forest). It also gives a typical minimum N, giving the agent clear signals on when and when not to use this tool.

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