Causal Decision Agent

Overview Schema Source Related Servers Discussions Admin

discos

Read-only

Estimates the causal effect of a treatment on an outcome distribution using distributional synthetic controls, comparing treated unit against a weighted donor pool and running placebo tests for inference.

Instructions

Distributional Synthetic Controls (Gunsilius 2023). Validation: validated evidence tier (known-truth, reference, external-parity, or Monte Carlo artifact). 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
`seed`	No	Random seed (currently unused; reserved for bootstrap extensions).
`time`	Yes	Time period column name.
`unit`	Yes	Unit identifier column name.
`alpha`	No	Significance level for confidence intervals.
`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	``'mixture'``: constrained (omega >= 0, Sigmaomega = 1) -- minimises the L2-Wasserstein distance between quantile functions. ``'quantile'``: unconstrained quantile-on-quantile regression.	mixture
`outcome`	Yes	Outcome variable column name.
`placebo`	No	Run in-space placebo permutation tests for inference.
`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.
`n_quantiles`	No	Number of quantile grid points on (0, 1).
`data_columns`	No	Optional column projection. Parquet/Feather/Stata loaders honour this for fast partial reads.
`treated_unit`	Yes	Value in unit that identifies the treated unit.
`data_sample_n`	No	Optional uniform random subsample size (seed=0, deterministic) — useful on huge panels.
`treatment_time`	Yes	First period of treatment (inclusive).

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

A4/5.0

Behavior4/5

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

Annotations already mark the operation as read-only, and the description adds meaningful behavioral context: SUTVA/no-anticipation assumptions, data preconditions, and failure modes with suggested remedies. It does not describe return behavior, but this is partly covered by the output schema and schema parameter descriptions.

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 dense but well-structured with labeled sections (Validation, Assumptions, Pre-conditions, Failure modes, Alternatives, Typical minimum N), making it scannable despite its length. Each section adds relevant information without redundancy.

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 complex econometric tool, this description covers assumptions, preconditions, failure modes, alternatives, and minimum sample size. Combined with the rich parameter schema and output schema, the agent has enough context to decide when and how to invoke the tool correctly.

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 parameters are already well documented. The description adds method-level context (convex donor combination, pre-treatment window, placebo inference) but does not meaningfully elaborate on specific parameter semantics beyond what the schema already provides.

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

Purpose4/5

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

Clearly identifies the method (Distributional Synthetic Controls, Gunsilius 2023) and communicates its intent through assumptions and preconditions. It does not use an explicit verb phrase like 'estimates the distributional treatment effect' and does not distinguish itself from closely related siblings such as discos_plot or discos_test.

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 alternatives (sp.sdid, sp.augsynth, sp.gsynth, sp.callaway_santanna), explicit preconditions, and failure-mode guidance such as switching to bias-corrected estimators when pre-treatment RMSPE is large. It lacks a clear when-not-to-use statement that would make routing among alternatives unambiguous.

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

Deploy Server