agentleFS
Fixed-SeedFixed-Seed/game-changer/AGENTS.md

game-changer

AGENTS.md8 starsChanged 35 days agoMITView on GitHub

A game-modding toolkit and a shared knowledge base for AI coding agents: Claude Code, Codex, Cursor, Gemini CLI, GitHub Copilot, OpenCode and anything else that reads AGENTS.md. When someone opens an agent in this repo, they almost always want to mod a game, or to learn how a game was modded.

The file

We split this AGENTS.md into the rules it gives an agent

9 rules. About 0 are common, 5 are distinctive, and 4 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈5

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 4

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Very specific:

    At the end, share what you learned: write a field note (bin/fsgc kb new ...) and, once your human agrees, open a PR (bin/fsgc kb pr <note> …

    line 12
  2. Very specific:

    Put it on PATH: export PATH="$PWD/bin:$PATH", or install it anywhere with uv tool install git+https://github.com/Fixed-Seed/game-changer.

    line 17
  3. Very specific:

    Every group has --help:

    line 19
  4. Very specific:

    No MCP? fsgc fixedseed does the same from the shell.

    line 37
  5. Distinctive:

    Never touch online clients protected by anti-cheat.

    line 46
  6. Distinctive:

    Never write cheats against other players (aimbots, ESP, speed hacks).

    line 47
  7. Distinctive:

    Never bypass anti-cheat, DRM or ownership checks.

    line 48
  8. Distinctive:

    What you ship: never commit or publish game files, extracted assets or decompiled code.

    line 52
  9. Distinctive:

    Wording: keep skills and docs agent-neutral ("the agent", not a product name) except in sections about one specific agent.

    line 68

≈ “Common” is an estimate: at least 1 in 5 of the 40,246 AGENTS.md files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
AGENTS.md

game-changer

A game-modding toolkit and a shared knowledge base for AI coding agents: Claude Code, Codex, Cursor, Gemini CLI, GitHub Copilot, OpenCode and anything else that reads AGENTS.md. When someone opens an agent in this repo, they almost always want to mod a game, or to learn how a game was modded.

Start here

  1. Read skills/mod-any-game/SKILL.md and follow its loop: intake → recon → route → lab (backups) → source of truth → vertical slice → assets → verify in the real game → showcase → publish.
  2. Search the knowledge base first: bin/fsgc kb search "<game or engine>". Other agents may already have written down the exact versions, routes and gotchas (knowledge/INDEX.md).
  3. At the end, share what you learned: write a field note (bin/fsgc kb new ...) and, once your human agrees, open a PR (bin/fsgc kb pr <note> --yes). See knowledge/README.md and CONTRIBUTING.md.

Tools

  • bin/fsgc is the Python CLI, and it sets itself up with uv.
    • Put it on PATH: export PATH="$PWD/bin:$PATH", or install it anywhere with uv tool install git+https://github.com/Fixed-Seed/game-changer.
    • Every group has --help:
      • scan: installed games, engine, anti-cheat, loaders, saves, routes
      • passthrough: two games at once. plan picks the host, each engine's hooks, the link and milestones; peer stands in for either side of the link
      • fixedseed (alias fs): sprites, textures, PBR, 3D, remeshes, rigs, motion, SVG, SFX, music, voice, video via the FixedSeed API (https://run.fixedseed.com, key in FIXEDSEED_KEY)
      • sprite / render3d: art → engine-ready frames
      • win: launch, screenshot, input, record on Windows (also from WSL)
      • video: contact sheets and EDL showcase edits
      • backup: snapshot and restore saves
      • publish: pre-release lint
      • kb: the knowledge base
  • FixedSeed MCP server: ships in this repo as bin/fsgc fixedseed mcp (stdio, stdlib only; code in game_changer/fixedseed_mcp.py). It reads FIXEDSEED_KEY from the environment. It also has a passthrough tool and prompt (the same plan as fsgc passthrough plan), which need no key.
    • It's pre-configured per agent: .mcp.json (Claude Code), .codex/config.toml (Codex), .cursor/mcp.json (Cursor), .vscode/mcp.json (VS Code / Copilot), gemini-extension.json (Gemini CLI).
    • No MCP? fsgc fixedseed does the same from the shell.
  • Skills (skills/*/SKILL.md, Agent Skills format) are also linked where each agent looks for them: .agents/skills (Codex and others), .claude/skills, .gemini/skills, .github/skills.
  • Engine playbooks: skills/mod-any-game/references/engines/.
  • Worked examples: examples/terraria-tmodloader, examples/aoe2-de-civ, examples/minecraft-gta5-passthrough.

Rules (full reasoning in skills/mod-any-game/references/safety.md)

  • What you can mod: any game the user owns: single-player, multiplayer, or servers the user hosts.
    • Never touch online clients protected by anti-cheat.
    • Never write cheats against other players (aimbots, ESP, speed hacks).
    • Never bypass anti-cheat, DRM or ownership checks.
  • Saves: fsgc backup saves before modded launches.
  • Spending: price a plan first (fsgc fixedseed estimate, MCP estimate) and tell the user the range; ask before spending more than about $5. Suggest a separate, spend-limited FixedSeed key for the agent.
  • What you ship: never commit or publish game files, extracted assets or decompiled code. Keep decompiles outside the repo.
  • Processes: kill by exact PID (fsgc win kill), never by name pattern.
  • Ask first before:
    • driving the user's mouse and keyboard, and never focus a game with an online mode while they're typing;
    • installing loaders into game folders or changing the registry;
    • publishing anything, PRs included.
  • Keep a journal: a MODLOG.md in the mod's working folder. It becomes your field note at the end.

Working on the toolkit itself

  • Python: 3.10+, deps Pillow, numpy and PyYAML (uv handles them via bin/fsgc). Code lives in game_changer/, one module per CLI group, each with register(sub) and a docstring that doubles as --help.
  • Windows tools: game_changer/ps1/*.ps1 embed C# 5 (Windows PowerShell 5.1's compiler): no string interpolation, no out var, no expression-bodied members.
  • Tests: uv run --with pytest pytest -q tests. CI also runs fsgc kb check --index and fsgc publish check ..
  • Wording: keep skills and docs agent-neutral ("the agent", not a product name) except in sections about one specific agent.

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.