DEV Community: sampo shen

commit-story-generator — turn git commits into adventure stories + engineering release notes

sampo shen — Wed, 11 Feb 2026 12:11:08 +0000

This is a submission for the GitHub Copilot CLI Challenge

What I Built

I built commit-story-generator, a cross-platform CLI that turns git commit history into two copy/paste-ready narratives:

Adventure style (release story with chapters)
Engineering style (PR/release summary with risk/validation/rollback)

It can also emit a categorized changelog (feat/fix/refactor/docs/test/chore) and basic stats.

Repo: https://github.com/aisamposhen/commit-story-generator

Demo

Reproducible 90-second demo (no git needed):

pip install -e .[test]

story template
story generate --fixture --style adventure --format md --dry-run
story release-notes --fixture --format md --dry-run
story changelog --fixture --format md --dry-run
story stats --fixture --format md --dry-run

My Experience with GitHub Copilot CLI

I used Copilot CLI as a terminal-first pair programmer to iterate module-by-module:

A robust git log reader (using record/field separators to avoid breaking on newlines)
Parser + classifier (feat/fix/refactor/docs/test/chore)
Story templates (adventure + engineering) with traceable short SHAs
Markdown + JSON rendering
Pytest coverage and CI matrix (Python 3.11 + 3.14)

The biggest win was tight feedback loops: prompt → run tests → refine behavior and UX.

terminal-escape-room — a tiny CLI escape room game (Python + Typer)

sampo shen — Tue, 10 Feb 2026 22:39:53 +0000

This is a submission for the GitHub Copilot CLI Challenge

What I Built

I built terminal-escape-room, a cross-platform CLI text escape-room game.

You progress through rooms defined in JSON, request up to 2 hints per room, submit answers, earn points (with hint penalties), and your progress is persisted locally to a .escape_state.json save file.

Tech stack:

Python 3.11+
Typer (CLI)
Rich (output)
Pytest (tests)

Repo: https://github.com/aisamposhen/terminal-escape-room

Demo

Quick run:

python3 -m venv .venv
source .venv/bin/activate
pip install -e .[test]

escape start
escape status
escape hint
escape answer "kernel"
escape reset

90-second demo outline:

escape start
escape status
escape hint
escape answer "kernel"
escape answer "^"
escape answer "git switch -c feature"

My Experience with GitHub Copilot CLI

I used GitHub Copilot CLI as a terminal-first pair programmer to iterate quickly on:

Typer command structure (start/status/hint/answer/reset)
A small engine layer (room progression + hint limits)
A stable JSON persistence schema (started_at/finished_at)
Scoring rules + tests

What worked well was keeping prompts very specific (inputs/outputs, edge cases, and file boundaries), then immediately locking changes in with pytest.

Example workflow that felt effective:

Ask Copilot CLI for a first-pass implementation (focused on one module at a time)
Run tests
Ask Copilot CLI to generate/adjust tests to cover an edge case
Refactor once the behavior is pinned

If you want to try it locally, everything is self-contained and requires no external services.

PatchPilot: Run Copilot diff apply verify (Copilot CLI Challenge)

sampo shen — Tue, 10 Feb 2026 12:17:38 +0000

This is my submission for the GitHub Copilot CLI Challenge.

What I built

I built PatchPilot, a small command-line tool that closes the loop:

run a failing build/test command
ask GitHub Copilot CLI to make the smallest fix inside the repo
capture a real git diff
re-run the same command to verify it’s green
write a short report you can share/review

Repo: https://github.com/aisamposhen/patchpilot

Why it’s useful

A lot of “AI in terminal” demos stop at suggestions. PatchPilot is opinionated: no celebration until the command actually passes again.

This makes it great for:

quickly unblocking on small test/build failures
creating minimal, reviewable patches
teaching the habit of verifying fixes

Demo (90 seconds)

# install (dev)
python3 -m venv .venv
source .venv/bin/activate
pip install -e .

# run the intentionally broken sample project
cd sample_projects/python_pytest_broken
pip install -r requirements.txt
patchpilot run --show-prompts --branch patchpilot/demo -- pytest -q

# view the generated report
sed -n '1,140p' PATCHPILOT_REPORT.md

How I used Copilot CLI (explicit)

PatchPilot calls Copilot CLI non-interactively via gh copilot.

The key design choice is that Copilot edits the repo in-place (with permission restricted to the repo directory), and PatchPilot then captures the patch using git diff. That keeps the result audit-friendly.

Example (simplified):

gh copilot -- --add-dir <repo> --allow-all-tools -p "...make pytest -q pass..."

UX / safety

Requires a clean worktree before running (so diffs are trustworthy)
Creates a new branch (defaults to patchpilot/fix)
Never pushes automatically
Stops after one attempt (keeps it predictable)

What’s next

If I had more time:

add one retry loop (bounded)
optional gh pr create to open a PR with the report
support multiple languages/framework templates

Thanks for reading!

cdoctor: a tiny CMake C 'code doctor' built with GitHub Copilot CLI

sampo shen — Tue, 03 Feb 2026 13:59:18 +0000

This is a submission for the GitHub Copilot CLI Challenge.

What I Built

I built cdoctor, a small CLI + CMake-based sample project designed to make C/C++ build/debug loops faster.

The idea: run one command, get:

a clean build attempt
a readable summary of warnings/errors
a proposed minimal patch (diff)
a re-run to confirm the fix

Demo

Repo: https://github.com/raungeguieia/cdoctor

Quick demo (copy/paste):

cmake -S . -B build
cmake --build build
./build/cdoctor --help
./build/cdoctor scan sample_project/src/warn.c
ctest --test-dir build --output-on-failure

My Experience with GitHub Copilot CLI

I used Copilot CLI in the terminal to:

summarize compiler output into actionable steps
propose patches for common C/CMake problems
generate small functions/tests quickly (then I reviewed and adjusted)

Example prompts I used:

“Given this CMake/clang error, propose the smallest fix as a git diff.”
“Add directory recursion to scan all .c/.h files on macOS+Linux.”
“Write a CTest smoke test for the CLI.”

Notes for Judges

This is intentionally small and easy to run.
The sample project includes an intentional warning to prove the pipeline works.