Causal Decision Agent

Overview Schema Source Related Servers Discussions Admin

propensity_score

Read-only

Estimate propensity scores to balance treated and control groups, enabling causal effect estimation from observational data while addressing confounding.

Instructions

Estimate propensity scores P(D=1|X). Assumptions: Unconfoundedness: treatment is as-good-as-random given the measured covariates; Overlap / common support: every unit has a non-degenerate probability of each treatment; The covariate set blocks all back-door paths. Pre-conditions: Pre-treatment covariates measured for treated and control units; A binary (or low-cardinality) treatment indicator; Sufficient covariate overlap between treatment arms. Failure modes: Poor overlap -- extreme propensity scores or few acceptable matches -> Trim or restrict to the common-support region and report the discarded units; Covariate imbalance remains after matching/weighting -> Re-specify the balancing model (CBPS, entropy balancing) and re-check standardized mean differences. Alternatives: sp.propensity_score, sp.cbps, sp.ebalance, sp.dml. Typical minimum N: 200.

Input Schema

TableJSON Schema

Name	Required	Description	Default
`detail`	No	Payload depth: 'minimal' (~150 tokens) for sub-step calls where only the point estimate is needed; 'standard' (~1K tokens) for diagnostics + coefficient table; 'agent' (~2K tokens, default) adds violations / next_steps / suggested_functions so the LLM can plan its next call without another round-trip.	agent
`method`	No	Estimation method. ``'logit'`` uses IRLS (no sklearn needed). ``'probit'`` uses scipy.optimize. ``'gbm'`` tries sklearn GradientBoostingClassifier, falling back to logit with interactions.	logit
`trimming`	No	If ``'crump'``, apply Crump et al. (2009) trimming after estimation. Trimmed observations receive ``NaN`` scores.
`as_handle`	No	If true, cache the fitted result on the server and return result_id + result_uri alongside the JSON payload so a subsequent tools/call can chain without re-running.
`data_path`	Yes	Absolute path or URL to a data file. Supported: .csv / .tsv / .txt (delimited), .parquet / .pq, .feather / .arrow, .xlsx / .xls, .dta (Stata), .json / .jsonl. Schemes: file://, s3://, gs://, https://.
`result_id`	No	Optional handle to a previously-fitted result (returned by an earlier call when as_handle=true). Tools that operate on a fitted object accept this in place of re-supplying data_path + columns.
`treatment`	Yes	Name of binary treatment column (0/1).
`covariates`	Yes	Covariate column names.
`data_columns`	No	Optional column projection. Parquet/Feather/Stata loaders honour this for fast partial reads.
`data_sample_n`	No	Optional uniform random subsample size (seed=0, deterministic) — useful on huge panels.

Output Schema

TableJSON Schema

Name	Required	Description	Default
No arguments

Schema Changelog

Changes observed during successful MCP inspections.

First observedv0.1.0Sep 15, 2026

TDQS

A4.4/5.0

Behavior5/5

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

Beyond the readOnlyHint annotation, the description discloses the core statistical assumptions (unconfoundedness, overlap, back-door blocking) and gives actionable failure-mode guidance such as trimming to common support and re-specifying the balancing model. This is substantial context for correct use and does not contradict 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.

Conciseness4/5

Is the description appropriately sized, front-loaded, and free of redundancy?

The description is longer than average, but the labeled sections—Assumptions, Pre-conditions, Failure modes, Alternatives, and Typical minimum N—make it scannable and each section carries decision-relevant information. The 'Typical minimum N' is advisory rather than a hard requirement, so it is slightly less crisp than a fully maximal definition.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Completeness5/5

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

For a high-complexity causal-inference estimator, the description covers validity conditions, failure modes, and alternative tools, while the output schema handles return-value details. An agent has enough information to decide whether to call this tool and how to interpret poor results.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Parameters3/5

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

The input schema already has 100% description coverage and documents all 10 parameters, including enums, defaults, and field roles. The description's P(D=1|X) notation clarifies the conceptual role of treatment and covariates but adds little beyond the schema. Baseline 3 is appropriate.

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Purpose5/5

Does the description clearly state what the tool does and how it differs from similar tools?

Opens with a precise mathematical statement of the estimand, P(D=1|X), and clearly identifies the tool as a propensity-score estimator. It adds assumptions, preconditions, and alternatives that distinguish it from many related siblings. There is no vagueness or tautology.

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Usage Guidelines4/5

Does the description explain when to use this tool, when not to, or what alternatives exist?

Provides explicit pre-conditions for valid use and names concrete alternatives such as sp.cbps, sp.ebalance, and sp.dml. However, it does not explicitly state selection criteria among those alternatives, leaving the agent to infer when this tool should be preferred rather than another.

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

Deploy Server