callaway_santanna

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

Annotations already declare readOnlyHint=true, so the description's burden is reduced, but it adds value by disclosing specific limitations: clustervars not supported with bstrap=False, analytic standard errors ignoring cluster dependence, and the necessity of multiplier bootstrap. This goes beyond the annotations and helps the agent anticipate constraints. The vague 'Validation: certified evidence' line is a minor detractor.

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, compact sentence that front-loads the core method and purpose, then tacks on necessary limitations and requirements. It avoids redundancy with the schema for most parameters. The 'Validation: certified evidence' phrase is filler but not harmful. Overall, it's appropriately sized and efficient.

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 tool's complexity (12 parameters, DID estimation) and the presence of an output schema, the description covers the essential method, required inputs (g, i, t, y implied), and known limitations. It does not explain the detail parameter's role in payload depth, but that is in the schema. The description, combined with a rich schema and annotations, is adequate for an agent to call the tool correctly.

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 75%, which is below the 80% threshold, so the description should compensate for undocumented parameters. It does explain the g parameter's role (cohort column) and the estimator options (dr/ipw/reg), but it offers no additional meaning for other parameters like detail, as_handle, or control_group. The schema already covers most of g and estimator, so the description adds marginal value beyond 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?

The description clearly identifies the tool as a Callaway-Sant'Anna staggered DID estimator computing group-time ATTs with three estimator options (doubly-robust, IPW, regression-adjusted). It states the method's advantage over TWFE, which helps disambiguate from many DID siblings. However, it does not explicitly name any sibling (e.g., staggered_cs, stacked_did) to differentiate from, so it's not a perfect 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?

The description implies usage when TWFE fails or heterogeneous treatment effects are present ('Robust to heterogeneous treatment effects where TWFE fails'), but it does not explicitly state when not to use it or name alternative tools. The 'Requires a cohort column g' is a prerequisite but not a usage guideline. This falls under implied usage rather than explicit routing.

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