cbps

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

Annotations only declare readOnlyHint=true and openWorldHint=false, so the description adds substantial behavioral context: unconfoundedness/overlap assumptions, pre-conditions, failure modes, and suggested remedies. It also notes a typical minimum sample size of 200 and mentions trimming/re-specification. It does not contradict the read-only hint—estimation is a read operation—and adds value beyond the 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?

The description is dense but organized with labeled sections (Assumptions, Pre-conditions, Failure modes, Alternatives, Typical minimum N), making it scannable. The opening sentence identifies the estimator immediately. Some content, such as 'Validation: certified parity evidence,' is vague and does not clearly earn its place, preventing 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?

For a causal estimator with 16 parameters and no nested objects, the description covers the essential usage context: identification assumptions, data pre-conditions, failure modes with remedies, alternatives, and a sample size guideline. Since an output schema exists, return-value details are not required. Missing is an explicit statement of what the tool returns (e.g., ATE estimate, balance diagnostics), but the parameter names and schema largely cover that.

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 every parameter, including enum meanings (detail, variant, estimand). The tool description adds general context (overlap, trimming, alternatives) but does not add parameter-level semantics beyond what the schema provides. Baseline 3 is appropriate because the schema carries the burden.

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 the exact estimator ('Covariate-Balancing Propensity Score estimator') and cites the method (Imai-Ratkovic 2014), so an agent can identify what this tool implements. However, it never states the primary verb/outcome explicitly—it does not say it estimates an average treatment effect, propensity scores, or balance weights, leaving that to inference from the parameter names (treat, y, estimand). This is clear but not maximally explicit.

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 concrete pre-conditions (pre-treatment covariates, binary treatment, sufficient overlap) and failure-mode guidance (trim to common support, re-specify balancing model), which tells an agent when the tool is appropriate and what to do when assumptions fail. It lists alternatives (propensity_score, ebalance, dml) but does not give explicit 'use this instead when X' routing, so it stops short of a 5.

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