DEV Community: george

Solstice Leap: leaping across the longest day with binary tape

george — Sun, 14 Jun 2026 16:03:34 +0000

This is a submission for the June Solstice Game Jam.

What I Built

I built Solstice Leap, a 3D browser arcade game where a glowing sun leaps across floating daylight platforms before the day fades out.

The game starts with a very simple input: hold to charge, release to jump. From there, each run becomes a small timing puzzle about reading distance, managing daylight, and choosing how far to push the streak. The player lands on normal platforms, prism boosts, shadow tiles, moving ledges, and binary tape platforms that write bits into a Turing-inspired protocol.

Live game: https://huanglianggit.github.io/solstice-leap/

Source code: https://github.com/huanglianggit/solstice-leap

Video demo: Watch the 34-second MP4 demo

Video Demo

The demo shows the opening run, distance-based jumping, binary tape platforms, prism boosts, and the first protocol transition from Dawn Tape into Prism Relay.

Direct video link: solstice-leap-demo-short.mp4

How It Fits the Theme

The solstice theme is built into the core loop rather than only the visuals.

The player is literally carrying a sun across the longest day. Daylight acts as the run timer, and every jump either preserves, restores, or drains that light. Prism platforms restore daylight, shadow tiles make survival harder, and the color palette shifts around warm sunlit platforms against a darker evening world.

I wanted the game to feel like a tiny ritual of keeping the day alive: charge, leap, land, recover light, and keep moving before sunset catches up.

Gameplay

Solstice Leap is intentionally readable at first, but it adds pressure through platform variety and protocol goals:

Daylight platforms are the baseline landing targets.
Prism boosts restore daylight and reward momentum.
Binary tape platforms write the current protocol bit.
Shadow tiles are smaller and drain daylight.
Moving ledges appear later and add timing pressure.

The player scores more for clean center landings and for building streaks. If a jump is undercharged but the sun still lands on the original platform, the run continues. The player only loses when the sun actually falls between platforms or daylight runs out.

How I Built It

The game is a static web project built with:

HTML
CSS
JavaScript
Three.js
Web Audio API

The 3D scene is rendered with Three.js. I used simple geometry, custom materials, canvas-generated textures, lighting, shadows, particle trails, and platform animations instead of external art assets.

The jump system is based on charge duration. A held input maps to travel distance, and the landing check compares the final position against the target platform and the original platform. That made the game feel more fair: a tiny hop that stays on the same pillar is not treated as a miss.

The protocol system generates binary platforms in sequence. Landing on a binary platform writes the next bit into the active tape. Completing a tape advances the run into the next named protocol, starting with Dawn Tape and then moving into Prism Relay, Turing Noon, and Longest Day.

I also added responsive HUD layout, local best-score storage, touch-friendly controls, and a GitHub Pages deployment so the game can be played directly in the browser.

Prize Category

I am submitting this for Best Ode to Alan Turing.

The game uses protocol names, binary tape, deterministic bit targets, and tape-writing platforms as a playful arcade interpretation of computation. The player is not only jumping across platforms. They are also helping a tiny sun-powered machine write and complete binary sequences before daylight runs out.

What I Learned

The most interesting design challenge was balancing a one-button arcade game with enough structure to feel fresh. A pure jumping game became much more interesting once the platforms also carried meaning: daylight recovery, risk, binary progress, and protocol transitions.

I also learned that small fairness details matter a lot. The jump should only fail when the player truly misses the platform, not when they make a harmless hop in place. Fixing that made the controls feel more trustworthy.

What's Next

If I continue developing Solstice Leap, I would like to add:

A sound toggle and richer landing sounds
More protocol variations
A few visual transitions between dawn, noon, and sunset
A leaderboard or shareable run summary
Optional mobile haptics for jump charging and landing

Thanks for checking it out.

AI Agent Skills Are Everywhere. But Which Ones Are Safe to Install?

george — Sat, 13 Jun 2026 09:04:04 +0000

AI agents are becoming less like single chatbots and more like extensible runtimes.

Instead of using a bare model, people are adding skills: coding helpers, research workflows, browser automation, MCP integrations, content generation tools, project-specific instructions, and more.

That is powerful.

But it also introduces a new trust problem.

When you install a third-party skill, you may be giving your agent new instructions, new tool access patterns, new dependencies, and sometimes new scripts. A skill can look helpful in its README while still containing risky behavior.

For example:

reading .env, SSH keys, browser cookies, or API tokens
running curl ... | sh
uploading data to a webhook
using destructive shell commands
telling the agent to ignore higher-priority instructions
including huge prompt files that waste tokens
depending on unpinned or unnecessary packages

As skill ecosystems grow, we need better pre-install checks.

That is why I built SkillPreflight.

What is SkillPreflight?

SkillPreflight is an open-source CLI and GitHub Action that scans AI agent skills before installation.

It performs static analysis and generates a 100-point scorecard covering:

Category	What it checks
Security	Dangerous commands, secret access, exfiltration, prompt injection, remote script execution
Permission restraint	Overbroad activation, unnecessary shell/network/file access
Token efficiency	Oversized skill files, repeated content, poor progressive disclosure
Lightweight footprint	File count, total size, dependencies, large assets
Maintainability	README, license, metadata, examples, documentation hygiene
Reliability	Tests, fixtures, deterministic workflow, error handling
Compatibility	Hardcoded paths, OS-specific assumptions, fragile shell usage

Usage

You can run it without installing anything globally:

npx skill-preflight@latest scan ./my-skill

You can also scan a GitHub repository:

npx skill-preflight@latest scan https://github.com/owner/repo.git

Generate JSON:

npx skill-preflight@latest scan ./my-skill --format json --out report.json

Generate SARIF for GitHub Code Scanning:

npx skill-preflight@latest scan ./my-skill --format sarif --out skill-preflight.sarif

GitHub Action

SkillPreflight is also available as a GitHub Action:

name: SkillPreflight

on: [pull_request, push]

jobs:
  scan:
    runs-on: ubuntu-latest
    steps:
      - uses: actions/checkout@v4
      - uses: agent-contracts/skill-preflight@v1
        with:
          target: "."
          fail-below: "70"

Safety model

SkillPreflight does not execute scripts from the skill being scanned.

It only reads files and performs static analysis. The goal is not to guarantee that a skill is perfectly safe, but to surface obvious red flags before users install it.

Why this matters

Browsers have extension permission prompts.

npm has package audits.

Containers have image scanners.

AI agent skills should have pre-install checks too.

As the AI agent ecosystem grows, users need a simple way to ask:

What does this skill actually do, and is it safe enough to install?

SkillPreflight is an early attempt at that.

GitHub:
https://github.com/agent-contracts/skill-preflight

npm:
https://www.npmjs.com/package/skill-preflight

GitHub Marketplace:
https://github.com/marketplace/actions/skillpreflight

I’d love feedback from people building or installing AI agent skills. What checks should be added next?