Skip to main content
Glama
davidesg

ART MCP Server

by davidesg

guided_identification

Guide Box-Jenkins model identification through sequential decision nodes: choose lambda, differencing, seasonality, and ARMA orders one step at a time.

Instructions

Sequential identification — ONE decision node per call.

DECISION TREE — call in this sequence, one at a time:

Call 1 lam=-1 (default) → Box-Cox scatter. Decide λ. WAIT for user.

Call 2 lam=X d=-1 (default) → Series(λ) + ACF/PACF at level d=0. ¿Trend? → next call with d=1. ¿No trend? → next call with d=0, D confirmed. Support: unit_root_analysis available if needed. WAIT for user.

Call 3 lam=X d= D=-1 → Series(λ) differenced d times + ACF/PACF + HAC seasonality. Seasonal? + B1 (deterministic seasonality: harmonics, D=0): Confirm d and D=0, then: a) confirm_and_estimate(m00: harmonics only, p=0, q=0) b) preliminary_outlier_scan on m00 residuals c) [cycle: add steps → re-estimate → scan] until clean d) Call 4 with pre_path=<mNN.pre> (ARMA on clean residuals) Seasonal? + B2 (stochastic seasonality: seasonal differencing, D=1): → Call 4 with lam, d, D=1 (ARMA+P+Q on ∇∇_s series) ¿No seasonality? → D=0, no harmonics, Call 4 directly. WAIT for user to confirm d and D.

Call 4 lam=X d= D= [pre_path=<.pre>] B1 path (D=0, pre_path given): → ACF/PACF of clean model RESIDUALS from pre_path. PACF cuts → AR(p). ACF cuts → MA(q). Also: mean significant? (μ̄/SE > 2) → estimate_mu=True B2 path (D=1, no pre_path): → ACF/PACF of ∇^d ∇_s y(λ). Also check lags s,2s,3s for seasonal P and Q. B1 no-outliers (D=0, no pre_path): → ACF/PACF of ∇^d y(λ) directly. WAIT for user to confirm p, q (and P, Q if D=1).

Parameters

inp_path : path to series .inp file (all calls) lam : Box-Cox lambda (-1 = not yet decided → Call 1) d : differencing order (-1 = not yet decided → Call 2) D : seasonal differencing (-1 = not yet decided → Call 3) pre_path : path to fitted .pre (Call 4, B1): ARMA identified on its residuals instead of the raw transformed series.

Input Schema

TableJSON Schema
NameRequiredDescriptionDefault
DNo
dNo
lamNo
inp_pathYes
pre_pathNo
Behavior4/5

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

With no annotations provided, the description carries the full burden. It discloses the tool's interactive nature through repeated 'WAIT for user,' its stateful flow via lam/d/D/pre_path parameters, and the B1/B2 branching logic. This goes beyond a simple action description, though it still omits explicit detail about return formats or side effects.

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

Conciseness3/5

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

The description is lengthy and includes a complete decision tree, but it is well-structured with sections and front-loaded with the core concept. Some repetition (e.g., 'WAIT for user') could be trimmed to improve conciseness, but each paragraph adds necessary procedural detail.

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?

The description covers the full decision tree, parameter semantics, and references to related tools (unit_root_analysis, confirm_and_estimate, etc.), making it quite complete for a complex interactive tool. However, it doesn't describe the output schema or return values explicitly, relying on the user to know what 'Box-Cox scatter' or 'ACF/PACF' entail.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Parameters5/5

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

The description includes a dedicated Parameters section explaining each parameter's meaning and default sentinel values (-1), and how pre_path is used in Call 4 B1. This fully compensates for the 0% schema coverage, providing rich semantics that the schema lacks.

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 'Sequential identification — ONE decision node per call' and outlines a multi-step decision tree, which distinguishes it as a guided, interactive identification tool. However, it doesn't explicitly contrast with sibling tools like identification_analysis, so the purpose is clear but not fully differentiated.

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 a detailed call sequence and mentions 'Support: unit_root_analysis available if needed,' giving some context on when to use this tool. However, it lacks explicit when-to-use vs. alternatives like boxcox_analysis or seasonal_analysis, and doesn't say when not to use this tool.

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

Install Server

Other Tools

Latest Blog Posts

MCP directory API

We provide all the information about MCP servers via our MCP API.

curl -X GET 'https://glama.ai/api/mcp/v1/servers/davidesg/art-python'

If you have feedback or need assistance with the MCP directory API, please join our Discord server