GitHub Repository

An LLM-powered agentic forecaster that uses sktime-mcp to autonomously discover, build, and evaluate optimal forecasting pipelines.

0 starsPython

An agentic loop for time-series forecasting using sktime and MCP

Name: An agentic loop for time-series forecasting using sktime and MCP
Availability: InStock
Author: silkpeak

by silkpeak·Apr 6, 2026·1 point·0 comments

Visit Project View on HN

AI Analysis

●●SolidBig BrainDark Horse

MCP proxy prevents context-window bloat better than naive LLM forecasting.

Strengths

•ReAct loop iteratively selects models instead of single-shot LLM guessing
•Mock LLM backends enable testing without API keys or costs
•Works with any LLM provider via simple callable interface

Weaknesses

•Agentic forecasting is trendy—AutoML tools already handle model selection
•Zero stars and forks suggests untested in production workflows

Post Description

Hi HN,

I was frustrated that feeding large time-series datasets directly into LLM prompts is highly inefficient, expensive, and prone to context-window hallucinations. To solve this, I built sktime-agentic-forecastor as an open-source proof-of-concept.

Instead of a single-shot prompt where the LLM blindly guesses a model, this uses a true ReAct (Reasoning + Acting) loop. It leverages the newly introduced Model Context Protocol (MCP) to proxy data safely into an ephemeral environment.

You can check out the source code, architecture diagrams, and run the quickstart examples here: https://github.com/amruth6002/sktime-agentic-forecastor

I'd love feedback on the overall architecture