AI Agent Use Cases
Views —

AI Agent Use Cases

Added today

44All 189

Prompt to Prepare a Shopping Cart from a Recipe Photo

The author reports that Grok Bot receives a recipe photo, organizes the ingredients and quantities, and creates a shopping list by excluding ingredients already available at home.

Grok Bot
Life & Hobbies

1–20 of 189

Work & Research

Finding Exoplanet Candidates in Public Telescope Data

The author, who is not an astronomer, reports analyzing TESS data over nine days in a single Claude Code session to find exoplanet candidates.

Claude Code

How it works

  1. The author planned the project and gathered reference materials with other models, but all execution was handled by Opus 5.5 in a single Claude Code cloud session.
  2. They continued the work over nine days, making use of the Pro plan's usage limit resets.
  3. The final analysis was reviewed separately using the models listed in the initial plan, and the author registered predictions for future transit times in advance.

Results

The author reports finding a planet candidate 218 light-years away, about 3.75 times the size of Earth, with four observed transits in TESS data from 2019 to 2026. They also released a draft paper, data, and code.

Notes

This is a candidate, not a confirmed planet. To be added to the official list, a peer-reviewed paper is required, which the author is preparing.

Source

Reddit Original post ↗

Example prompt · Claude Code

Please help conduct an analysis to find exoplanet candidates in public space telescope data.

Input: The target star or list to analyze [e.g., TIC number], data source [e.g., TESS public light curves], relevant prior articles or papers [links], available computing environment and time [e.g., cloud session, several days]. First, check the official download path and usage conditions for the data, and search public catalogs to see if this star already has known planets or candidates.

Write up an analysis plan as a document and get my approval before proceeding. Follow these steps: clean the light curve, search for periodic signals, fit transit signals, and check for false positives (eclipsing binaries, noise, contamination from nearby stars). At each stage, save the code, plots, and reasoning. If the work continues over several days, leave progress notes so it can be resumed later.

When results are available, summarize the candidate's size, period, number of transits, and uncertainties, and predict future transit times. Clearly label the result as a candidate, not a confirmed discovery. Do not submit to official catalogs or journals, but explain the necessary procedures.

Deliverables: analysis code, plots, summary report, prediction table, and a list of remaining validation steps. First, run the same procedure on a star with a known planet to check that the results match published values.