Research

At CoreStory, we're conducting deep research to understand and shape AI's role in modern software development. Explore our latest research, including our Guiding Agents Mid-Task study and The Plausibility Floor, at frontier.corestory.ai.
Colorful cartoon landscape featuring Snoopy sitting on grass near a yellow doghouse, a multicolored tree, a rainbow-striped bench, and vibrant trees under a sunny sky.
COMPLETE

CHARLIE BROWN

Comparative analysis of LLMs vs. human ability to elicit unexpressed constraints for software requirements.
See details
Cartoon cat in a spacesuit joyfully floating in a colorful cosmic scene with planets, stars, and vibrant clouds.
COMPLETE

GUMDROP

Measuring and improving GitHub Copilot's code suggestion by expanding contextual awareness through neighboring tabs.
See details
Colorful cartoon space scene with a rocket, planets with cute faces including one with rings, stars, and a pink skull-shaped planet.
COMPLETE

COLUMBIA

Vision paper on a future SDLC with AI at its core.
See details
Colorful abstract illustration of planets, stars, and interconnected orbs in space with a purple background.
COMPLETE

Guiding Agents Mid-Task

A response-scoped guidance channel was read and acted on when it carried state unavailable elsewhere, tested on synthetic tool surfaces.
See details
Colorful 3D letter V with swirling patterns and glowing orbs in a vibrant cosmic background.
COMPLETE

THE PLAUSIBILITY FLOOR

A blind agent with no code access still scored 22.6–32.9/100 on a code-comprehension rubric, setting an evaluation baseline.
See details

Future Research

Our next studies are already underway. Follow them as they publish at frontier.corestory.ai.
Visit Frontier

A Word From Our Research Team

We’re really interested in how AI is going to impact Human Computer Interaction. More specifically, we’re interested in how AI will impact the software development lifecycle (SDLC).

As you may (or may not) know, finding people willing to help in research studies is the hardest part. We thank you deeply for your participation and willingness to help us ask the big questions, then answer them with real data.