Skip to main content

UE Knowledge Base

Offline semantic search over Unreal Engine development knowledge — a curated Chinese-language knowledge corpus (29 topics: GAS, animation, AI, networking, UMG, Niagara, PCG, ...) plus a local RAG pipeline that indexes it with BAAI/bge-small-zh-v1.5 embeddings into ChromaDB.

Zero API cost. Fully offline after a one-time model download. Built for Chinese-speaking UE developers, and ready to plug into any AI agent, IDE, or CLI workflow.

knowledge/  (29 topics, ~84 markdown docs)
    │  ue-kb build          chunk + embed (BGE small, local)
    ▼
.chroma_db/  (vector index)
    │  ue-kb query "GAS cooldown"
    ▼
top-k semantic hits with source + heading

Features

  • 📚 Curated corpus: hand-written UE development docs in Chinese, covering Gameplay Ability System, character movement, animation, physics/collision, AI navigation, networking/replication, UMG/Slate, Niagara, Mass Entity, State Trees, PCG/procedural generation, materials/rendering, module build system, editor tools, and more.
  • 🧠 Local semantic search: BAAI/bge-small-zh-v1.5 (multilingual, ~100MB)
    • ChromaDB — no cloud APIs, no cost, works on a laptop CPU.
  • 🖥️ Simple CLI (ue-kb): build, query, info, download-model, plus --json output for agent integration.
  • 🔌 Extensible: index any extra UE docs with --source, or add engine source-indexing scripts under scripts/.

Quick start

# 1. Install
pip install -e .                 # or: pip install ue-knowledge-base

# 2. Download the embedding model once (~100MB)
ue-kb download-model
#    In China:  export HF_ENDPOINT=https://hf-mirror.com  then retry

# 3. Build the index
ue-kb build

# 4. Search
ue-kb query "GAS ability cooldown" --top-k 5
ue-kb query "角色移动 速度衰减" --json   # machine-readable for agents

Example

$ ue-kb query "GAS 技能冷却"
🔍 UE 知识库检索:GAS 技能冷却

[1] ue-gameplay-abilities/references/gas-input-integration.md › 问题 (匹配度: 21.1%)
    UE 项目同时使用 GAS (GameplayAbilitySystem) 和 Enhanced Input 时,容易陷入
    两个极端:- **全 GAS** → 所有输入走 GAS,但 WASD 轴输入不适合 GAS 的事件
    模型,且 CommitAbility 的 GC 延迟影响跳跃手感 ...
[2] ue-gameplay-abilities/references/gas-input-integration.md › Jump — GAS 即时技能 (匹配度: 14.2%)
    ...

Python API

from ue_knowledge.query import query

for hit in query("GAS cooldown", top_k=5):
    print(hit["source"], hit["heading"], hit["score"])

Extending the corpus

  1. Add a markdown file under knowledge/<topic>/ (use ##/### headings — the indexer chunks on heading boundaries).
  2. ue-kb build --force to rebuild.

For indexing Unreal Engine C++ header comments or Epic official docs, see scripts/index_engine_source.py and scripts/crawl_epic_docs.py (they expect local engine/UE paths — the extracted index data is generated locally and is not redistributed, out of respect for Epic's copyright).

License

MIT. The knowledge documents are original writing; no engine source code or verbatim Epic documentation is included.

Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

ue_knowledge_base-0.1.0.tar.gz (11.0 kB view details)

Uploaded Source

Built Distribution

If you're not sure about the file name format, learn more about wheel file names.

ue_knowledge_base-0.1.0-py3-none-any.whl (10.5 kB view details)

Uploaded Python 3

File details

Details for the file ue_knowledge_base-0.1.0.tar.gz.

File metadata

  • Download URL: ue_knowledge_base-0.1.0.tar.gz
  • Upload date:
  • Size: 11.0 kB
  • Tags: Source
  • Uploaded using Trusted Publishing? No
  • Uploaded via: twine/7.0.0 CPython/3.11.15

File hashes

Hashes for ue_knowledge_base-0.1.0.tar.gz
Algorithm Hash digest
SHA256 ab413fa19cc004c74042cc51fa11346dac67050db5790fd380ef7b72e18810e1
MD5 7f1b2dbee1d1a0df164909eb9df26ee8
BLAKE2b-256 28e323e743000e3f035f5c6042615996f19bdaf41ebdd2f9234ba070312e01eb

See more details on using hashes here.

File details

Details for the file ue_knowledge_base-0.1.0-py3-none-any.whl.

File metadata

File hashes

Hashes for ue_knowledge_base-0.1.0-py3-none-any.whl
Algorithm Hash digest
SHA256 3f5ec3eb9be3dc1c2ade2105479010258b38b16a956f5ea750c3e13856d815a2
MD5 6cb061fd630660ab4f857759bae9f1f0
BLAKE2b-256 6bf18cc15a292742e7bbd923cbdceabe4b168ec9ed3d926ab7264009e905f48f

See more details on using hashes here.

Supported by

AWS Cloud computing and Security Sponsor Datadog Monitoring Depot Continuous Integration Fastly CDN Google Download Analytics Sentry Error logging StatusPage Status page