Real Skill packageSource verifiedClawHub registry

Local MiniGPT

从零实现的本地小型 GPT(Python + PyTorch,字符级分词,默认约 150 万参数,CPU 可训练)。当用户想在本地训练/微调一个自己的小型语言模型、学习 Transformer/GPT 的从零实现、用 Python 写一个"本地AI大模型"演示、或需要用自有语料做本地文本风格模仿与续写(不追求真实大模型质量)时使用。附带:零 Key 网页搜索工具、检索增强开卷对照、代码能力考试、MiniViT 手写数字识别。典型触发:训练一个小模型、从零实现 GPT、字符级语言模型、用我的语料训练模型、mini GPT、nanoGPT 式实现。不适用于:直接运行现成开源大模型(Qwen/LLaMA 等请用 transformers 或 Ollama)、生产级 NLP 任务。

Identity and source

Publisher attributionyifanQQMCregistry owner unverified by skillvetai
Functional categorySoftware Developmentautomatically inferred · 56% rule confidence
Package forminstruction with code19 recorded files
Canonical sourceClawHub registryclawhub:yifanqqmc:local-minigpt
Open canonical source ↗

Platform declarations

These states come from the source or distribution context. None of the entries below are SkillVetAI compatibility test results.

OpenClawnative officialProvenance: registry distribution

Independent structural checks

These checks parse the fixed package against dated platform rules. They do not execute the Skill or verify task behavior.

Claude Codepasses structure
Checker 0.1.0 · agent-skills-2026-08-13+claude-code-docs-2026-08-13 · 10/18/2026.claude/skills/local-minigpt

Runtime, accounts, dependencies, permissions, network behavior and task quality remain untested.

OpenAI Codexpasses structure
Checker 0.1.0 · agent-skills-2026-08-13+codex-docs-2026-08-13 · 10/18/2026.agents/skills/local-minigpt

Runtime, accounts, dependencies, permissions, network behavior and task quality remain untested.

OpenClawpasses structure
Checker 0.1.0 · agent-skills-2026-08-13+openclaw-docs-2026-08-13 · 10/18/2026skills/local-minigpt

Runtime, accounts, dependencies, permissions, network behavior and task quality remain untested.

Installation and inspection

This command is recorded from the source ecosystem and resolves the registry's latest release. The fixed release shown on this page should be inspected before adoption.

clawhub install @yifanqqmc/local-minigpt
clawhub inspect @yifanqqmc/local-minigpt --version 1.0.0

Security evidence

SkillVetAI static result: high signal

This automated, non-executing scan is bound to this release hash. It is not a safety certification and may contain false positives or false negatives.

Status
completed
Coverage
full text content
Files
19 / 19 inspected as text
Checked
10/18/2026, 4:29:09 AM
Scanner
0.1.3
Policy
1.0.3
7 automated findings
highDynamic code or shell execution is presentscripts/coding_test.py:92 · confidence 78%eval(
highDynamic code or shell execution is presentscripts/generate.py:50 · confidence 78%eval(
highDynamic code or shell execution is presentscripts/model.py:215 · confidence 78%eval(
highDynamic code or shell execution is presentscripts/search_demo.py:73 · confidence 78%eval(
highDynamic code or shell execution is presentscripts/train.py:82 · confidence 78%eval(
highDynamic code or shell execution is presentscripts/vision/predict.py:60 · confidence 78%eval(
highDynamic code or shell execution is presentscripts/vision/train_vit.py:90 · confidence 78%eval(
7 High/Critical review queue entries
STATIC_DYNAMIC_CODE_EXECUTIONpending
STATIC_DYNAMIC_CODE_EXECUTIONpending
STATIC_DYNAMIC_CODE_EXECUTIONpending
STATIC_DYNAMIC_CODE_EXECUTIONpending
STATIC_DYNAMIC_CODE_EXECUTIONpending
STATIC_DYNAMIC_CODE_EXECUTIONpending
STATIC_DYNAMIC_CODE_EXECUTIONpending
Open human review queue →
4 inferred permission indicators
  • network access — automatically inferred
  • filesystem read — automatically inferred
  • filesystem write — automatically inferred
  • browser control — automatically inferred
9 dependency and API indicators
  • api: clawhub.ai
  • api: download.pytorch.org
  • api: github.com
  • api: html.duckduckgo.com
  • api: ossci-datasets.s3.amazonaws.com
  • api: raw.githubusercontent.com
  • api: storage.googleapis.com
  • api: www.baidu.com
  • api: www.bing.com
External clawhub result: clean

This is registry-supplied evidence for the recorded release, not an independent SkillVetAI scan. Check the canonical source for the full report, scanner versions, scope, and current moderation state.

Evidence checked
10/18/2026, 3:06:50 AM
Release binding
Matches this record
  • vt: clean
  • skillspector: suspicious
  • llm: clean

Recorded files

The catalog stores hashes and an inventory summary for change detection. It does not republish the package contents.

Package content hashsha256:59c47986beb332cc18b38adfd63d47a9c88a5a5dae9ffeda45ac32b374c214c1
Show up to 19 recorded paths
  • data/demo.txt
  • scripts/build_code_corpus.py
  • scripts/coding_test.py
  • scripts/fetch_corpus.py
  • scripts/generate.py
  • scripts/model.py
  • scripts/search_demo.py
  • scripts/test_model.py
  • scripts/test_tokenizer.py
  • scripts/tokenizer.py
  • scripts/train.py
  • scripts/vision/data.py
  • scripts/vision/make_samples.py
  • scripts/vision/predict.py
  • scripts/vision/train_vit.py
  • scripts/vision/vit.py
  • scripts/web_search.py
  • skill-card.md
  • SKILL.md

Source changelog

1.0.0 首个版本:从零实现的本地 GPT(模型/分词/训练/生成)+ 零 Key 网页搜索 + 检索增强开卷对照 + 编码能力考试 + MiniViT 手写数字识别,附单元测试与中文文档