Causal Decision Agent

Overview Schema Source Related Servers Discussions Admin

causal_rl_benchmark

Read-only

Generate synthetic causal-RL benchmark datasets from logged trajectories to evaluate offline policies under ignorability, positivity, and Markov dynamics. Use offline-safe methods when behavior-policy coverage is poor.

Instructions

Generate a synthetic causal-RL benchmark dataset. Assumptions: Sequential ignorability: no unobserved confounders of actions and outcomes; Positivity: the behaviour policy explores all evaluated actions; The environment satisfies the assumed (Markov) dynamics. Pre-conditions: Logged trajectories (states, actions, rewards) from a known or estimable behaviour policy. Failure modes: Poor behaviour-policy coverage -- the target policy queries unseen state-action regions -> Use offline-safe / pessimistic methods and report effective sample size of the importance weights. Alternatives: sp.offline_safe_policy, sp.policy_value. Typical minimum N: 1000.

Input Schema

TableJSON Schema

Name	Required	Description	Default
`name`	No	'confounded_targeting', 'confounded_routing'}	confounded_bandit
`seed`	No	Random seed for reproducible stochastic steps.
`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
`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`	No	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_episodes`	No	Number of episodes.
`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.
`confounding_strength`	No	Magnitude of unmeasured confounding U -> (action, reward).

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

A3.7/5.0

Behavior4/5

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

The description discloses key assumptions (sequential ignorability, positivity, Markov dynamics), pre-conditions, failure modes, and typical sample sizes. These go well beyond the sparse annotations (only readOnlyHint), giving the agent a clear picture of the tool's behavior and limitations. It 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 a single dense paragraph but efficiently packs useful information: purpose, assumptions, pre-conditions, failure modes, and alternatives. It is front-loaded with the primary purpose and avoids fluff. A more structured format (bullets) might improve scannability, but the current length is justified.

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

Completeness4/5

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

With an output schema present, the description does not need to detail return values. It covers critical contextual aspects—assumptions, failure modes, and typical N—that help an agent decide when and how to use the tool. It also hints at chaining via as_handle (indirectly) but that is covered by the schema.

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 coverage is 100% with each parameter described, so the baseline is 3. The description adds general guidance like 'Typical minimum N: 1000' but does not elaborate on individual parameters beyond what the schema already provides, so it neither enhances nor detracts from parameter understanding.

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?

The description clearly states a specific verb ('Generate') and resource ('synthetic causal-RL benchmark dataset'), making the tool's purpose obvious. It distinguishes itself from the many analysis-focused siblings by being a data-generation tool, though it does not explicitly contrast with other benchmark generators like recommend_benchmark or verify_benchmark.

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

Usage Guidelines3/5

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

The description provides pre-conditions (logged trajectories), typical minimum N, and failure modes with alternative tools (sp.offline_safe_policy, sp.policy_value). However, it does not explicitly state when to use this tool over other dataset generators, and the alternatives are framed as responses to failure rather than as direct alternative choices for the same task.

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

Deploy Server