Causal Decision Agent

Overview Schema Source Related Servers Discussions Admin

causal_dqn

Read-only

Learn a confounding-robust Q-function from logged trajectories for offline policy learning under unobserved confounding.

Instructions

Causal deep Q-network (Li, Zhang, Bareinboim 2025, arXiv:2510.21110) for offline policy learning under unobserved confounding. Learns a confounding-robust Q-function via bootstrap data augmentation. 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
`state`	Yes	State column(s)
`action`	Yes	action parameter (str).
`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
`n_iter`	No	Fitted-Q iterations
`reward`	Yes	reward parameter (str).
`discount`	No	Discount factor
`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.
`next_state`	Yes	Next-state column(s)
`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

A4.3/5.0

Behavior4/5

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

Annotations already declare readOnlyHint=true and openWorldHint=false, establishing the tool as non-mutating. The description adds valuable behavioral context: lists assumptions (sequential ignorability, positivity, Markov dynamics), pre-conditions (logged trajectories from a known behaviour policy), and failure modes. It does not contradict annotations and goes beyond them with actionable caveats.

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: starts with the core purpose, then assumptions, pre-conditions, failure modes, alternatives, and typical N. Each sentence contributes useful guidance. Slightly long but efficiently organized with no filler.

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?

For a complex tool with 12 parameters and no nested objects, the description covers assumptions, failure modes, alternatives, and sample-size guidance. It does not explain return values, but an output schema exists. The description is complete enough for an agent to decide when to use it and what to expect at a high level.

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%, with each of the 12 parameters having a descriptive comment. The tool description adds no parameter-specific details beyond the schema, so it meets the baseline 3 for a fully documented schema. It does not repeat or override schema information.

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 description clearly states the tool's function: learns a confounding-robust Q-function via bootstrap data augmentation for offline policy learning under unobserved confounding. It explicitly distinguishes itself from alternatives like offline_safe_policy and policy_value, making its specific role unambiguous.

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

Usage Guidelines5/5

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

The description provides explicit usage guidance: it names alternatives (sp.offline_safe_policy, sp.policy_value), specifies failure modes (poor coverage) and recommends switching to offline-safe methods, and gives a typical minimum sample size (1000). This gives an agent clear decision rules for when to invoke this tool.

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

Deploy Server