Half-Life 2 rebuild deck
VERY HARD
A fixed accumulator and an honest mass table get you the gravity gun by Sunday; the game is fifteen hours of set pieces authored around it one week at a time.
You will get a good prototype. The product is what comes after.
The KINDA is the gravity gun. Mass-aware props plus a clamped hold force is a weekend, and what you end up holding is recognisably the thing Valve invented, which is not true of most games this size. The agent will put the physics step inside the render loop; it works perfectly on your machine and then a thrown radiator passes clean through an enemy on somebody's 144Hz display. Insist on the fixed accumulator and CCD before you tune anything. Spend the rest of your review on the mass table, because flat masses quietly delete the design — if a paint tin kills, nothing in the corridor means anything. What you will not have is fifteen hours of authored pacing, and that is most of the game.
- Difficulty
- VERY HARD · 4 of 5
- First playable
- A project
- The original
- 250 team-years
- Released
- 2004 · PC
- Genre
- FPS
Build one physics-driven corridor in the Half-Life 2 mould: a counterweight puzzle, then an ambush. three.js plus a rigid body library such as Rapier. One hitscan weapon, one gravity gun. Not the campaign. Requirements: - Fixed 60Hz physics accumulator decoupled from rendering, with interpolated draw, and continuous collision detection on any body moving further than half its own size in a step. Thrown props are the primary weapon here, and a discrete solver sends a fast radiator straight through an enemy without touching it. Verify the same throw kills at 30fps and 144fps. - Every prop carries an authored mass and density in its spawn data. A cinder block is not a paint tin. The gravity gun refuses to lift above a mass threshold, and anything below its fluid density floats. - Impact damage scales with kinetic energy at contact, so a heavy prop at speed kills and a light one bounces off. Never give props a flat damage value. - The puzzle is a plank on a fulcrum with a cage at each end, solvable only by moving mass, plus a water section where buoyancy does the lifting. - Four enemies with cover positions and burst fire, ragdolling on death with the killing impulse carried through. - A companion who follows and never blocks a doorway: push her out of the player's capsule rather than pathing around it. Everyone cheats here. Definition of done: I kill an enemy by launching a radiator, the same throw with a cardboard box does nothing, and both behave identically at any frame rate.
- Around fifteen hours of pacing during which the camera is never once taken away from you, which is a design achievement rather than an engineering one
- Hammer, the phoneme-driven facial animation, and the scripted-sequence tooling that lets NPCs act out a scene while you wander off
- Set pieces authored around physics one at a time — the seesaw, the ramp, the sand — each of which is somebody's week
- Alyx, who is the reason the middle third of the game works at all
Constraint-driven carry — held objects are pulled toward a hold point by a clamped damped spring, never parented to the camera.
Any carry, tractor beam, telekinesis or physics drag wants this. Parenting makes objects clip through walls and feel weightless; a force makes them fight you, clatter into things, and communicate their mass without a single UI element.
ScopeAn evening— build just this, not the game
Build a constraint-driven carry toy as a single self-contained HTML file, Canvas 2D, no build step. Requirements: - Eight circles of differing mass in a box: gravity, wall collisions, body-to-body impulses. The solver is not the point. - Click a body to hold it. Never set its position and never parent it to the cursor. Each step, apply a force of k times cursor minus position, minus c times velocity, clamped to a maximum that scales with mass. - Because that is a force and not a teleport, a held body still collides: drag it into a wall and it stops there, drag it through a pile and it shoves the pile. Releasing throws it at the spring velocity. - Break the hold if the body ends a step too far from the cursor; that is what stops you dragging things through geometry. Definition of done: a heavy body lags behind the cursor, a held body pressed into a wall stops dead, and raising the clamp until it punches through shows why the clamp is there.
Not a clone — Half-Life 2 re-themed and cut down to something you can finish. The angle is the important reel: it is what turns a project into a weekend. 5,880 ways to land.
Metro Endless
Half-Life 2, but an endless survival mode instead of levels, set in a superhero city. Rendered in 16-colour pixel art.
Build a browser game called "Metro Endless". The pitch in one line: a fps in the spirit of Half-Life 2, but an endless survival mode instead of levels, set in a superhero city. WHAT THIS IS Take the SHAPE of Half-Life 2 — its core loop, and whatever makes that loop good — and rebuild it as your own game in a different world at a smaller scope. Use your own names, art and audio, which you want anyway: a reskin of somebody else's game is less interesting than a new one that works the same way. THE ANGLE — this is the important constraint, honour it above everything else An endless survival mode instead of levels. One arena, escalating waves on a data-driven timeline, a run timer and a score. Replaces level design with a difficulty curve. This is what makes the project finishable. If a decision would push the scope back towards the original's, take the smaller option every time. THE THEME A superhero city. Powers that are verbs, a city built to be traversed with them, and collateral damage the player has to care about. The theme is not a coat of paint. Let it change what the mechanics mean — a reload, a health pack and a locked door should all be things that make sense in this world and nowhere else. LOOK 16-colour pixel art. A fixed 16-colour palette (start from a Lospec palette), a hard pixel grid, and no sub-pixel movement — snap all rendering to whole pixels. REQUIREMENTS - Runs in a browser with no build step. Canvas 2D unless the angle genuinely requires 3D, in which case three.js. - Fixed-timestep update loop with interpolated rendering, so it behaves the same at 60Hz and 144Hz. - Every tuning constant in one CONFIG object at the top, exposed as live sliders in a debug panel. You will find the feel by dragging those, not by prompting. - All content — levels, entities, balance numbers — in JSON or plain data files, never inline in the logic. - Audio generated with the Web Audio API rather than asset files. - Respect prefers-reduced-motion: keep the fades, drop the shake and the parallax. BUILD ORDER 1. The core loop with placeholder rectangles, at the reduced scope the angle demands. No theme, no art, no audio. Make it fun as rectangles first. 2. The angle's consequences. Cutting to bots, or to turn-based, or to one level changes the design rather than just shrinking it — find out how and rewrite step 1. 3. The theme, as content and data. 4. The look, applied last as a rendering layer over a game that already works. Start by writing a one-page plan: the core loop, what the angle removes and what that frees you to do properly, and what you are deliberately leaving out. Wait for me to approve it before writing code.
- Rapier — fast deterministic rigid body physics with a WASM build
- three.js — the rendering layer for the prototype
Half the time one of these already does what you wanted, and the other half it saves you a week of solving a problem someone documented in 1997.
One count per game. No account — it is tied to a salted hash of your IP, which is not stored in a form anyone can reverse. Every build adds 250 team-years to the counter on the homepage.
Then try these
Questions
How hard is Half-Life 2 to rebuild with an AI agent?
Yes — every game is buildable at some scope. The question is how hard, and Half-Life 2 is VERY HARD (4 of 5). You will get a good prototype. The product is what comes after. A fixed accumulator and an honest mass table get you the gravity gun by Sunday; the game is fifteen hours of set pieces authored around it one week at a time. Reckon on a project to something you can actually play — not to something that matches the 250 team-years Valve spent.
What makes Half-Life 2 hard to rebuild?
A fixed accumulator and an honest mass table get you the gravity gun by Sunday; the game is fifteen hours of set pieces authored around it one week at a time. That is why it sits at VERY HARD rather than a tier either side of it. If you want the idea without the project, the transferable part is constraint-driven carry — held objects are pulled toward a hold point by a clamped damped spring, never parented to the camera. — an evening of work.
How long does it take to build a Half-Life 2 clone?
Wrong question, and it is worth saying why. Nobody rebuilds Half-Life 2 — the original took Valve on the order of 250 team-years, released in 2004 on PC. What you can do is reach something playable and genuinely yours in about a project, and then keep going for as long as it stays interesting. Any figure that claims otherwise is measuring a prototype and calling it a game.
What do I lose by building my own Half-Life 2?
Around fifteen hours of pacing during which the camera is never once taken away from you, which is a design achievement rather than an engineering one. Hammer, the phoneme-driven facial animation, and the scripted-sequence tooling that lets NPCs act out a scene while you wander off. Set pieces authored around physics one at a time — the seesaw, the ramp, the sand — each of which is somebody's week. Those are the parts a prompt will not hand you.
Which AI coding agent should I use to build it?
Any of Claude Code, Codex or Cursor will handle this prompt. Paste it as your first message, let the agent scaffold the project, then iterate on feel — the second, third and tenth prompts are where a game actually gets good. Work in an empty folder or a fresh git branch so you can throw it away cheaply.
One email when new games go on the list. No other reason to email you, ever.