Causal Decision Agent

Overview Schema Source Related Servers Discussions Admin

scest

Read-only

Estimate synthetic control weights that reproduce a treated unit’s pre-treatment outcome path from donor units, enabling causal effect estimation after treatment.

Instructions

Estimate synthetic control weights. Assumptions: A convex (or regularized) combination of donor units reproduces the treated unit's pre-treatment outcome path; No interference: the treatment does not affect the donor units (SUTVA); No anticipation before the treatment date. Pre-conditions: Panel of one or more treated units plus an untreated donor pool, observed over time; Pre-treatment window long enough to fit donor weights (rule of thumb: more pre-periods than donors used); Outcome observed for every unit in every period. Failure modes: Large pre-treatment RMSPE -- the synthetic unit fails to track the treated unit before treatment -> Add donors / predictors, lengthen the pre-period, or use a bias-corrected estimator (sdid, augsynth); Placebo / permutation inference shows the estimate is not extreme relative to donors -> Report the placebo distribution honestly; the effect may not be distinguishable from noise. Alternatives: sp.sdid, sp.augsynth, sp.gsynth, sp.callaway_santanna. Typical minimum N: 15.

Input Schema

TableJSON Schema

Name	Required	Description	Default
`time`	Yes	Time period column name.
`unit`	Yes	Unit identifier column name.
`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
`outcome`	Yes	Outcome variable column name.
`w_constr`	No	Weight constraint: - ``'simplex'`` : w >= 0, sum(w) = 1 - ``'lasso'`` : L1-penalised (allows negative, non-summing) - ``'ridge'`` : L2-penalised - ``'ols'`` : ordinary least squares (unconstrained) - ``'ls'`` : least squares (same as 'ols')	simplex
`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.
`data_columns`	No	Optional column projection. Parquet/Feather/Stata loaders honour this for fast partial reads.
`lasso_lambda`	No	L1 penalty (used when ``w_constr='lasso'``).
`ridge_lambda`	No	L2 penalty (used when ``w_constr='ridge'``).
`treated_unit`	Yes	Identifier of the treated unit.
`data_sample_n`	No	Optional uniform random subsample size (seed=0, deterministic) — useful on huge panels.
`treatment_time`	Yes	First treatment period.

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

A4.5/5.0

Behavior5/5

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

Annotations already declare readOnlyHint=true, and the description adds substantial behavioral context: the convex/regularized weighting assumption, SUTVA/no-anticipation requirements, and failure modes such as large pre-treatment RMSPE and weak placebo inference. This goes well beyond what the annotations alone convey.

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Conciseness5/5

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

The description is dense and well-structured, with labeled sections for assumptions, pre-conditions, failure modes, and alternatives. Every sentence carries useful information, and the core purpose is front-loaded.

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?

Given a 100%-covered input schema and an output schema, the description supplies the remaining non-obvious context: when the method is valid, what can go wrong, how to respond to failure modes, and which alternatives exist. Nothing an agent needs to decide whether to invoke this tool is missing.

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?

Schema description coverage is 100%, so the schema already documents every parameter. The description only loosely maps to parameters (e.g., 'convex or regularized' hints at w_constr but never names it), so it adds little parameter-specific value beyond the schema baseline.

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?

The first line, 'Estimate synthetic control weights,' uses a specific verb and resource and immediately distinguishes this from generic causal estimators. The assumptions, pre-conditions, and named alternatives reinforce exactly what kind of method this is.

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?

The description gives explicit pre-conditions (treated units plus donor pool, sufficient pre-period, complete outcomes) and failure-mode remedies, including named alternatives like sdid and augsynth. However, it does not fully spell out when-not-to-use or give a crisp decision rule among all listed alternatives.

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

Deploy Server