Godot Save Load Systems logo

Godot Save Load Systems

Community
thedivergentai
godot-save-load-systems

Expert blueprint for save/load systems using JSON/binary serialization, PERSIST group pattern, versioning, and migration. Covers player progress, settings, game state persistence, and error recovery. Use when implementing save systems OR data persistence. Keywords save, load, JSON, FileAccess, user://, serialization, version migration, PERSIST group.

Overview

Publisherthedivergentai
RepositoryGD-Agentic-Skills
Skill namegodot-save-load-systems
Stars
782
Forks
48
Bundled files
6
LicenseLGPL-3.0
Links
  • Markdown instructions

    A SKILL.md file the model loads on demand, so it only costs tokens when a request actually matches.

  • Works with any LLM

    AI skills are plain Markdown, not provider-specific code, so this works with GPT, Claude, Gemini, Grok, or a local model.

  • 6 bundled files

    Scripts, templates, and references the model can read while it works. Files are read-only and never executed.

  • Open source

    Published by thedivergentai on GitHub. Read the source before you install it.

Installation

Install the Godot Save Load Systems AI skill in TypingMind to use it with any LLM, or drop it into another agent that reads SKILL.md.

1

Install in TypingMind

TypingMind installs a skill straight from its GitHub folder — it reads SKILL.md, bundles the resource files, and stores the result locally.

  1. Open the app and go to Plugins → Skills.
  2. Choose "Install from GitHub".
  3. Paste the skill folder URL below and confirm.
    https://github.com/thedivergentai/GD-Agentic-Skills/tree/main/skills/godot-save-load-systems
  4. Enable the skill in any chat where you want it available.
Plugins → Skills → Add skill → From GitHub URL, then paste the folder URL and press Continue.
2

Install in another agent

Any agent that reads the Agent Skills format can use this skill — copy the folder into that agent's skills directory.

Claude Code — .claude/skills
git clone --depth 1 https://github.com/thedivergentai/GD-Agentic-Skills.git /tmp/GD-Agentic-Skills
mkdir -p .claude/skills
cp -r /tmp/GD-Agentic-Skills/skills/godot-save-load-systems .claude/skills/godot-save-load-systems
Restart Claude Code after copying so it picks up the new skill.

Use it in TypingMind

Enable Godot Save Load Systems in any TypingMind chat and the model takes it from there. Its name and description sit in the system prompt, and the moment a request matches, the model loads the full instructions itself — you never invoke it by hand, and it costs no tokens until it is actually used.

The model loads Godot Save Load Systems on its own as soon as a request matches it.

Works with any AI model

AI skills are plain Markdown instructions rather than provider-specific code, so Godot Save Load Systems is not tied to the model it was written for. Install it once in TypingMind and use it with GPT-5, Claude, Gemini, Grok, DeepSeek, Mistral, Llama, or a local model you run yourself — all on your own API keys.

  • Loaded only when it is needed

    The system prompt carries just the name and description. The instructions are fetched on the first matching request, so an idle skill costs nothing.

  • Switch models mid-chat

    Because the skill is instructions rather than code, changing model does not break it — the next model reads the same SKILL.md.

Skill instructions

This is the SKILL.md content the model loads. Read it before installing — a skill is instructions your model will follow.

NEVER Do

  • NEVER save without a version field — When you update your game's data structure, old saves will break. Always include a "version": "1.0.0" field and implement migration logic.
  • NEVER use absolute OS paths — Hardcoding C:/Users/... will break on every other machine. Always use the user:// protocol, which Godot maps to the correct OS-specific app data folder.
  • NEVER attempt to save Node references directly — Nodes are objects, not raw data. Extract the necessary primitive data (positions, health, levels) into a Dictionary or Resource instead.
  • NEVER forget to close FileAccess handles — Leaving a file open can lead to handle leaks and save-file corruption. In Godot 4, files auto-close when the variable goes out of scope, but explicit close() is safer for long-running logic.
  • NEVER use JSON for very large binary data — Storing 10MB of texture data as Base64 in JSON is slow and bloats file size. Use binary store_var() or separate dedicated asset files.
  • NEVER trust loaded data without validation — Users can edit save files. Always use data.get("field", default_value) and validate that numbers are within expected ranges to prevent crashes.
  • NEVER trigger a save during high-frequency physics or animation updates — A crash mid-write will corrupt the file. Save only on explicit game events like entering a menu, finishing a level, or at a checkpoint.
  • NEVER modify a save Dictionary while iterating over its keys — Calling erase() or add() inside a loop over the same dictionary causes iteration errors. Use data.duplicate() to iterate safely.
  • NEVER store raw passwords or sensitive credentials in unencrypted JSON — If you have sensitive data, use FileAccess.open_encrypted_with_pass() to secure it.
  • NEVER use ResourceLoader.load() for massive scenes on the main thread — It causes a visible freeze. Use ResourceLoader.load_threaded_request() to load levels in the background.
  • NEVER rely on get_instance_id() for cross-session identification — These IDs are assigned at runtime and change every time the game restarts. Generate your own persistent String UUIDs for game objects.
  • NEVER forget to call duplicate(true) on a loaded Resource stats block — If multiple enemies load the same "goblin_stats.tres", they will all share the same health pool unless duplicated.
  • NEVER use the "allow_objects" flag in store_var/get_var for untrusted data — Setting this to true allows full object decoding, which is a major security risk for saves downloaded from the web.
  • NEVER use JSON for data requiring strict type preservation — JSON converts Vector3 to a string or dictionary. For strict data types, use var_to_bytes() or a binary format.
  • NEVER leave internal metadata (set_meta) in persistent dictionaries — This unnecessarily inflates save file size. Clean your dictionaries before serialization.

Available Scripts

MANDATORY: Read the script for the chosen format before writing SaveManager code.

save_load_patterns.gd

MANDATORY for JSON / binary / PERSIST collect — patterns default store_var(..., false).

save_migration_manager.gd

MANDATORY when any save has a version field that can lag the build.

save_system_encryption.gd

MANDATORY before encrypted slots — password from secure storage / user secret, never hardcoded in examples.

save_integrity_validator.gd

Rolling .bak + SHA-256 verify before trusting a slot; fall back to backup on mismatch.


Deep dive (load on demand)

MANDATORY for JSON/PERSIST walkthroughs, binary examples, gotchas, and elite encrypted paths — references/save-patterns-deep.md. Do not paste Step 1–3 Autoload tutorials into scenes.

Expert WHY (critical)

CAUTION: Baseline tutorials used store_var(data, true). Untrusted user:// saves must allow_objects=false — RCE risk on modded/workshop files.

  • Vectors in JSON — store {x,y,z} components; JSON does not round-trip Vector3 faithfully.
  • Rolling backup — crash mid-write corrupts primary; copy to .bak before overwrite (save_integrity_validator.gd).
  • When to save — menu/checkpoint/level complete only — never per physics frame.

Decision Tree: Pick a Persistence Shape

NeedFormatMANDATORY
Human-readable, small/medium progressJSON + versionsave_load_patterns.gd
Type-faithful Variants / larger blobsBinary store_var with allow_objects=falsesame
Typed Resource trees / inspector schemasResourceSaver / .tres/.resPeer godot-resource-data-patterns
Many scene nodes auto-collectPERSIST group + save()/load()save_load_patterns.gd
Schema evolvedMigrate then loadsave_migration_manager.gd
Anti-tamper / sensitive fieldsEncrypted FileAccesssave_system_encryption.gd

Do not paste Step 1–3 JSON Autoload tutorials here — implement from the scripts.

allow_objects Trust Boundary

Default always store_var(data, false) / get_var(false).

Caseallow_objectsRule
Player user:// saves, workshop mods, downloadsfalseNEVER true — RCE risk
Trusted local only (your own tooling, offline debug fixtures you control)true only if unavoidableDocument why; never ship as default; prefer Resources / Dictionaries of primitives

Encrypted elite paths still use false unless the payload is explicitly trusted-local and non-user-editable.

Golden Path (version → migrate → backup → atomic write)

  1. Version field on every save blob.
  2. Migrate via save_migration_manager.gd when versions differ.
  3. Backup existing file (DirAccess.copy_absolute to .bak) before overwrite.
  4. Write to temp then rename, or write-after-backup; validate open errors.
  5. Integrity optional: FileAccess.get_sha256 compare; fall back to backup on mismatch.
  6. Paths only user:// — never absolute OS paths.
  7. When to save — menu, checkpoint, level complete — never per physics frame.

Settings may use ConfigFile separately from run-progress JSON/binary.

Reference

Progressive disclosure: open Official Documentation links only when researching a specific API; load Related Skills when routing to a peer domain — do not preload the whole lattice.

Official Documentation

  • Saving games — Persist group serialization, JSON line format, and the canonical save/load loop this skill builds on.
  • File paths in Godot — user:// vs res:// mapping across OS app-data folders; never hardcode absolute paths.
  • Binary serialization API — store_var/get_var Variant encoding, type fidelity, and why allow_objects is unsafe for untrusted saves.
  • Background loading — ResourceLoader.load_threaded_request for hitch-free level/resource loads after a save restore.
  • File system — FileAccess/DirAccess workflow for existence checks, backups, and safe overwrite patterns.
  • Resources — Resource vs Dictionary persistence, duplicate(true), and when .tres/.res beats hand-rolled JSON.
  • Groups — SceneTree group membership used by the Persist/PERSIST auto-collect pattern.
  • FileAccess — Open modes, encrypted-with-pass AES helpers, SHA-256, compression flags, and buffer I/O.
  • JSON — stringify/parse/parse_string for human-readable saves and validation of parse errors.
  • ConfigFile — INI-style settings (user://settings.cfg) separate from full game-state saves.
  • ResourceSaver — Persist typed Resources/custom Resource trees when JSON type loss is unacceptable.
  • AESContext — Low-level AES block encrypt/decrypt used by custom compressed encrypted save pipelines.

Related Skills

Prerequisites
  • godot-project-foundations — ProjectSettings, Autoload registration, and user:// project identity must exist before a SaveManager can own paths.
  • godot-gdscript-mastery — Typed Dictionaries, Resources, and error-handling patterns for versioned serialize/deserialize code.
  • godot-autoload-architecture — SaveManager is almost always an Autoload; use this for singleton ownership, boot order, and scene-change survival.
Complements
  • godot-resource-data-patterns — Custom Resource schemas and .tres workflows that pair with ResourceSaver instead of flattening everything to JSON.
  • godot-scene-management — Threaded scene swaps and wipe/rebuild Persist nodes after load without leaking old world state.
  • godot-signal-architecture — game_saved / game_loaded event buses so UI and systems react without hard-wiring SaveManager.
  • godot-ui-containers — Settings menus that write ConfigFile/volume keys this skill persists separately from run progress.
  • godot-inventory-system — Item stacks and equipment dictionaries are the heaviest Persist payloads; share ID schemes with save migration.
  • godot-quest-system — Quest flags/stage IDs must round-trip through versioned saves without breaking journal UI.
  • godot-economy-system — Currency wallets and shop unlocks need the same version field and validation as player progress.
Downstream / consumers
Master
  • godot-master — Library router and mirrored module entry; open when discovering which Domain Skill owns persistence vs content systems.

Bundled files

The model reads these on demand while the skill is loaded. They are exposed as readable files and are never executed.

Frequently asked questions

What does the Godot Save Load Systems AI skill do?

Expert blueprint for save/load systems using JSON/binary serialization, PERSIST group pattern, versioning, and migration. Covers player progress, settings, game state persistence, and error recovery. Use when implementing save systems OR data persistence. Keywords save, load, JSON, FileAccess, user://, serialization, version migration, PERSIST group.

Why use Godot Save Load Systems on TypingMind?

Because you install it once and use it with any model. Godot Save Load Systems is plain Markdown rather than provider-specific code, so the same skill runs on GPT-5, Claude, Gemini, Grok, or a local model — and you can switch model mid-chat without it breaking. TypingMind runs on your own API keys, so you pay providers directly instead of a per-seat subscription, and your skills and chats stay in your own storage.

How do I install Godot Save Load Systems in TypingMind?

Open Plugins → Skills → Install from GitHub in TypingMind and paste https://github.com/thedivergentai/GD-Agentic-Skills/tree/main/skills/godot-save-load-systems. TypingMind reads its SKILL.md and bundles its files and installs it as a skill you can enable per chat.

Which AI models can use Godot Save Load Systems?

Any model you connect in TypingMind. AI skills are plain Markdown instructions rather than provider-specific code, so GPT, Claude, Gemini, Grok, and local models can all load this skill when a request matches it.

How many AI models can I use with Godot Save Load Systems?

As many as you like. As long as a model supports skills, you can use Godot Save Load Systems with it — GPT, Claude, Gemini, Grok, DeepSeek, Mistral, Llama and more — all on TypingMind with your own API keys.

Is the Godot Save Load Systems AI skill free?

Yes. It is published on GitHub by thedivergentai under the LGPL-3.0 license. You only pay your own AI provider for the tokens you use.

What are AI skills?

An AI skill is a reusable instruction bundle that teaches an AI model how to do one specific task. It follows the open Agent Skills format: a SKILL.md file with a name and description, plus any scripts, templates or reference files the model may need. The model reads the instructions only when your request matches the skill, so an installed skill costs nothing until it is used.

How are AI skills different from plugins or MCP servers?

A plugin or MCP server gives a model new tools to call — code that runs somewhere and returns a result. An AI skill gives the model knowledge and process instead: how to approach a task, which steps to follow, what good output looks like. Skills are plain Markdown, so they need no server, no API key and no runtime, and they work with any model.

View all

Set up your own AI workspace now

Get notified about new features and future giveaways by subscribing to our newsletter 👇