Causal Decision Agent

Overview Schema Source Related Servers Discussions Admin

notch

Read-only

Estimate behavioral elasticity from bunching at a policy notch by fitting counterfactual density and measuring excess mass at the threshold.

Instructions

Bunching at Notches estimator (Kleven & Waseem 2013). Validation: validated evidence tier (known-truth, reference, external-parity, or Monte Carlo artifact). Assumptions: The counterfactual density would be smooth through the threshold absent the policy; Excess mass at the threshold reflects the behavioural elasticity of interest; No other discontinuity coincides with the threshold. Pre-conditions: A behavioural choice variable (earnings, hours, ...) with a known kink or notch in the budget/choice set; A visible empirical density of the running variable around the threshold. Failure modes: Round-number heaping or a coincident policy contaminates the bunching mass -> Exclude heaping points, widen the excluded region, and test the counterfactual polynomial order. Alternatives: sp.rdrobust, sp.rkd. Typical minimum N: 500.

Input Schema

TableJSON Schema

Name	Required	Description	Default
`x`	Yes	Running variable name (e.g., 'income').
`seed`	No	Random seed for reproducibility.
`alpha`	No	Significance level.
`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_boot`	No	Number of bootstrap replications for standard errors.
`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.
`bin_width`	No	Bin width for the histogram.
`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.
`notch_size`	No	Size of the discontinuous jump (delta-tau). If provided, a structural elasticity is estimated.
`poly_order`	No	Polynomial order for the counterfactual distribution.
`notch_point`	Yes	Location of the notch in the running variable.
`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.
`exclude_range`	No	Range around the notch to exclude from the counterfactual fit. If None, defaults to (notch_point - 3bin_width, notch_point + 5bin_width).

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?

Annotations declare readOnlyHint=true, so the read-only nature is already known. The description adds meaningful behavioral caveats: smooth counterfactual assumption, excess mass interpretation, contamination from heaping/policy, and mitigation strategies. This enriches the agent's understanding of what the estimator actually does and its failure modes.

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?

Appropriately compact, using labeled sections (Validation, Assumptions, Pre-conditions, Failure modes, Alternatives, Typical minimum N) so key information is easy to scan. The opening phrase immediately names the estimator, making the purpose front-loaded.

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?

Covers validation tier, assumptions, preconditions, failure modes, alternatives, and minimum sample size. With the full schema and an output schema present, this is quite complete. The only notable gap is failure to mention sibling bunching tools, which leaves some ambiguity in tool selection.

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 covers 100% of parameters with per-parameter descriptions, so the description need not repeat them. It does not add any parameter-level semantics beyond what is already in schema, but it doesn't need to; baseline 3 applies.

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?

States clearly that it estimates bunching at notches, with a specific method reference (Kleven & Waseem 2013). However, it does not distinguish itself from sibling tools 'bunching' and 'general_bunching', which likely have overlapping purpose.

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?

Provides explicit pre-conditions (behavioral variable with known notch, visible density) and a typical minimum N of 500, so it signals when the input data is suitable. But the listed alternatives (sp.rdrobust, sp.rkd) are RD methods and are not accompanied by any when-to-choose criteria, and it never mentions the 'bunching' or 'general_bunching' siblings, leaving the agent without guidance on selecting among bunching tools.

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

Deploy Server