skills.rest
timothyfraser
timothyfraserCOMMUNITY
Explore all skills by this author →
Website𝕏@Timothy_Fraser1GitHub

data-science-agent-loop

Automates data-science loops for batched Ollama Cloud tool calls with JSONL audits.

3|21|Updated Sep 19, 2025
Search Tags:#automation#audit#batching#data-science#agent-loop#data-cleaning#ollama-cloud
One-click install
npx skills add https://github.com/timothyfraser/dsai --skill data-science-agent-loop-timothyfraser
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: data-science-agent-loop
Source: https://github.com/timothyfraser/dsai/tree/main/.cursor/skills/data-science-agent-loop
Command: npx skills add https://github.com/timothyfraser/dsai --skill data-science-agent-loop-timothyfraser

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Enables engineers to scaffold repeatable, disposable data-science loops that orchestrate short-lived R or Python scripts invoking Ollama Cloud tools, batch processing rows, optionally parallelizing calls, and generating per-run JSONL audits within a folder-local environment.

Core Features & Use Cases

  • Single-folder loop: one main script plus optional helpers (R: functions.R or Python: helpers.py), with a local .env configuration.
  • Batched tool calls to Ollama Cloud, optional parallel chunking, and structured JSONL audits to track results.
  • Use cases include data cleaning, enrichment, and dataset compilation where repeatable loops improve consistency and reproducibility.

Quick Start

Create a new folder, add a main script and optional helpers, configure a local .env, and run the loop to batch calls to Ollama Cloud.

Frequently Asked Questions about data-science-agent-loop

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I automate batched data cleaning and enrichment with Ollama Cloud?▼

Automating batched data cleaning involves orchestrating short-lived scripts that call Ollama Cloud tools in chunks. You can configure a folder-local environment with a main script, optional Python or R helpers, and a .env file to process rows in batches and log structured JSONL audits.

What is a short-lived data-science agent loop and when do I need one?▼

A short-lived data-science agent loop is a disposable, single-folder orchestration that runs scripts invoking Ollama Cloud tools for specific tasks. You need one for repeatable data cleaning, geospatial routing, or dataset compilation where consistency and per-run audits are required.

Can I use Python and R helpers in the same automated data-science loop?▼

Yes, the automated data-science loop supports an optional R helper file named functions.R or a Python helper named helpers.py. You include your chosen helper alongside the main script in the folder to provide custom functions for the batched tool calls.

How do I track results and audits when processing datasets in parallel with Ollama Cloud?▼

To track results and audits during parallel dataset processing, the loop generates structured JSONL audits per run. These logs capture the outcomes of batched Ollama Cloud tool calls, ensuring you can review results and maintain reproducibility across your data-science tasks.

Does this automated data-science agent loop require external dependencies?▼

No external dependencies are required to use the automated data-science agent loop. You simply create a new folder, add your main script and optional helpers, configure a local .env file with your Ollama Cloud credentials, and run the loop to start batching calls.

What is the best way to control batch sizes for geospatial routing tasks with Ollama Cloud?▼

The best way to control batch sizes for geospatial routing tasks is by configuring the ROWS_PER_BATCH setting. This parameter manages how many rows are sent in each chunk to Ollama Cloud, allowing you to balance processing speed and optional parallelism.

Related Skills

Explore More Data & Analytics Skills→
robomotioniorobomotionio
2

ollama

Connect a CLI tool to a local Ollama server for text generation, chat, and embeddings.

Official
Intermediate
patricio0312revpatricio0312rev
5

doc-to-vector-dataset-generator

Convert documents into chunked JSONL datasets for embeddings.

Community
Advanced
CallMeLuigiv2CallMeLuigiv2

ollama

Educates developers on using Ollama's local LLM runtime and API endpoints.

Community
Advanced
vvnu0vvnu0

data-science-agent-loop

Batch data, dispatch tool calls to Ollama Cloud, and collect JSONL audits.

Community
Advanced