Idea Sprint: Stop Brainstorming, Start Building

Where this fits

You're staring at an empty repo, wondering if the crypto idea you've been chewing on for weeks is worth a single line of code. idea-sprint replaces that loop of self-doubt with a structured interview that ends in a decision — build, pivot, or drop — before you waste a sprint.

Why agents benefit

• Runs a blunt, one-question-at-a-time interview to extract edge, constraint, and wedge — no flattery, no filler
• Applies a necessity gate to kill ideas that solve problems nobody actually has
• Scores 3 candidates against explicit criteria and outputs a go/no-go decision, not a menu of options
• Persists context to .claude/context/idea.md so downstream skills (pitch-deck, hackathon) pick up where validation left off
• Resumes from prior state instead of re-interviewing, making it usable across multiple agent sessions

Practical scenarios

Solo dev with a vague idea

"I want to build something on Solana" — idea-sprint forces you to articulate your edge and constraints, then surfaces 3 concrete candidates you can actually ship with your skill set.

Pre-hackathon validation

You have 48 hours and a team of 3. Run idea-sprint on your concept before committing the weekend. The necessity gate tells you if it's worth the caffeine.

Portfolio of half-started projects

You've got 5 abandoned repos. Run idea-sprint on each concept to kill the weak ones and double down on the one with a real wedge.

Add it to your agent workflow

Install with a single command:

npx skills add solanabr/solana-ai-kit --skill idea-sprint

Then invoke it naturally:

> Validate my idea: a Telegram bot that lets users swap tokens without leaving the chat.

The agent runs the interview, scores your idea, and writes .claude/context/idea.md with the decision and open risks. For more context on how this skill fits the broader agent-skill ecosystem, see the idea-sprint skill page.

Compare before adopting

Before you adopt idea-sprint, compare it against plain prompt engineering (which lacks the structured interview and scoring), or manual frameworks like Y Combinator's startup school. Check the repo's commit freshness, whether the author responds to issues, and whether the skill's output format matches what your downstream tools expect.

Why it is worth tracking

idea-sprint comes from the solana-ai-kit repo (97 stars, MIT license, adapted from sendaifun/solana-new) — young but with real momentum in the Solana builder community. The telemetry-removed, context-persisting design signals a thoughtful author. Track it if you're building crypto agents and want a validation layer that actually produces decisions.