dml_diagnostics

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

Beyond the readOnlyHint and openWorldHint annotations, the description discloses statistical assumptions (unconfoundedness, overlap, Neyman-orthogonal scores, cross-fitting, nuisance convergence rates) and failure modes (propensity near 0/1 inflates variance/bias). This adds substantial behavioral context about what the report validates and what can go wrong.

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 front-loaded with the core action and then organizes assumptions, preconditions, failure modes, alternatives, and sample-size guidance into clearly labeled segments. It is dense but each sentence contributes substantive context; no 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 diagnostics tool with 8 parameters, an output schema, and complete schema coverage, the description covers the surrounding context well: assumptions, data requirements, failure handling, alternatives, and minimum sample size. Nothing essential for correct invocation 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 100%, so the baseline is 3. The description adds real value by specifying what the 'result' parameter must contain (post-fit residuals, propensity diagnostics for IRM), which goes beyond the schema description. Other parameters are not individually enriched, but the schema already documents them adequately.

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 opens with a specific verb and resource: 'Build a DMLDiagnostics report from a DML CausalResult.' This clearly states what the tool does and what input it consumes, and it distinguishes itself by listing alternatives (sp.tmle, sp.auto_cate, sp.causal_forest). It is not a tautology because it adds the source object type and report deliverable.

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 conditions for use: preconditions (covariates, treatment, outcome; instrument for PLIV/IIVM), failure-mode handling with a pointer to dml_sensitivity for overlap violations, and a 'Typical minimum N: 500' guideline. However, the three alternatives (sp.tmle, sp.auto_cate, sp.causal_forest) are named without explicit when-to-choose conditions, so guidance is clear but not exhaustive.

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