ghc-prompting-for-completions

Teaches steering GitHub Copilot inline completions using comments, signatures, and naming.

2|Updated Jun 1, 2026
One-click install
npx skills add https://github.com/jay-steenbergen/MSSAMentorAgent --skill ghc-prompting-for-completions-jay-steenbergen
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: ghc-prompting-for-completions
Source: https://github.com/jay-steenbergen/MSSAMentorAgent/tree/main/.github/skills/tracks/github-copilot/ghc-prompting-for-completions
Command: npx skills add https://github.com/jay-steenbergen/MSSAMentorAgent --skill ghc-prompting-for-completions-jay-steenbergen

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? Learners often accept whatever GitHub Copilot suggests without understanding that what they type before the suggestion controls its quality, leading to generic or wrong code and frustration with AI-assisted coding. ## Core Features & Use Cases - Three Prompt Levers: Structured drills teaching comment-as-intent, typed-signature-as-shape, and function-name-as-direction to steer inline completions. - Context Window Training: Hands-on exercises demonstrating the open-tabs trick, tab hygiene, and recognizing when Copilot is informed versus guessing. - Five-Function Gauntlet: A capstone exercise where learners build 5 working Python functions (date parsing, chunking, email masking, retry logic, grouping) by prompting deliberately and verifying output. - Use Case: A coding bootcamp student who just learned Copilot basics uses this project to move from passively accepting suggestions to actively controlling them, cutting the iterations needed to get correct code. ## Quick Start Ask the mentor to start the ghc-prompting-for-completions project so we can practice steering Copilot inline completions with comments, signatures, and naming.

Frequently Asked Questions about ghc-prompting-for-completions

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

FAQPage Schema
How do I get better GitHub Copilot inline suggestions?

Write a descriptive comment above the function, use typed signatures with precise parameter names, and choose a verb-accurate function name. These three levers give Copilot the context it needs to produce targeted suggestions instead of generic ones.

How does GitHub Copilot use open tabs for context?

Copilot reads the current file plus other open editor tabs, not your whole repository. Opening a relevant file like a models definition before writing code that uses it makes suggestions aware of actual field names and types.

What does Ctrl+Enter do in GitHub Copilot?

Ctrl+Enter opens the Copilot completions panel showing multiple alternative suggestions instead of just the first inline ghost text. Use it when the initial inline suggestion looks wrong so you can pick a better candidate.

Do type hints improve Copilot suggestion quality?

Yes, type hints are one of Copilot's primary signals. A signature like filter_older_than(users: Iterable[dict], min_age: int) -> list[dict] produces far more accurate completions than an untyped def filter(items, x).

When should I use Copilot Chat instead of inline completions?

Inline completions are the highest-volume interaction and best for writing new code line by line. Copilot Chat is better for debugging, explanations, and multi-step tasks, which is covered in the follow-up chat-driven debugging project.