match

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

The description is exceptionally transparent about behavioral traits beyond the annotations. It discloses: order-dependence of greedy matching with a quantified >5x spread, exact reproducibility conditions (m_order='data'/'closest' reproduce MatchIt), the StatsPAI-specific 'farthest' rule, bias_correction convention differences with quantified ~0.1% deviation, exactness of the uncorrected estimate and Abadie-Imbens SE, memory complexity O(n_treated x n_control), practical scale limits, and the unconfoundedness/overlap assumptions. The readOnlyHint=true annotation is consistent with a matching estimator that computes estimates without mutating data. This is far beyond what annotations provide.

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 and information-rich, with clear section labels (Validation, Known limitations, Cost, Assumptions). It front-loads the core purpose and then provides critical caveats. It is longer than ideal, but every sentence carries substantive information: quantified effect sizes, exact reproducibility rules, memory complexity, and statistical assumptions. The structure helps an agent scan for the relevant caveat. It loses one point for length and for the truncated 'Assumptions' section that cuts off mid-sentence ('tr...'), which is a structural flaw.

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 (19 parameters, 7 enums, 15 methods, 6 SE methods), the description is remarkably complete. It covers the key decision-relevant facts: what the tool does, when results are exact vs approximate, how to reproduce reference implementations, memory constraints, and statistical assumptions. The output schema exists, so return values need not be described. The only minor gap is the truncated assumptions sentence, but the core assumptions (unconfoundedness, overlap) are already stated. For a tool this complex, the description provides an agent with enough to select it and avoid known pitfalls.

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 baseline is 3. The description adds meaningful semantics beyond the schema: it explains the material impact of m_order (>5x spread), clarifies that 'caliper' is not a method but a parameter, explains the memory implications of the dense distance matrix, and gives the statistical meaning of se_method options with measured coverage properties. It also explains the bias_correction convention difference. This goes beyond the schema's per-parameter descriptions, though the schema itself is already rich. The description does not enumerate every parameter, but the schema covers them, so the combination is strong.

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+resource: 'Propensity score and covariate matching for treatment effect estimation.' It names the tool's domain (matching for causal treatment effects) and its core function. It distinguishes itself from siblings like psm, psmatch2, propensity_score, and optimal_match by covering the full matching family (nearest, kernel, radius, cem, etc.) and by being the general-purpose matching entry point. The description also includes validation and known-limitations context that makes the tool's identity unmistakable.

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 gives strong usage context: it states the tool is for treatment effect estimation via matching, and the 'Known limitations' section tells the agent when to be cautious (order-dependence, memory cost, bias-correction conventions). It also gives practical guidance like 'use a caliper or coarser blocking beyond that' for large data. However, it does not explicitly name sibling alternatives or state when to choose this tool over psm, psmatch2, propensity_score, or optimal_match. The guidance is implied by the tool's comprehensive scope rather than stated as explicit routing rules.

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