streamlit-core

仓库创建 2026年7月5日最近提交 22 天前SkillHot 收录 22 天前

▸ 精选理由

解决常见的重跑/状态丢失问题，是 Streamlit 开发必读。

这个 Skill 做什么

说明 Streamlit 的执行模型、session_state、缓存与 rerun 原理，便于调试。

帮你把 Streamlit 程序里那些莫名其妙的重跑、状态丢失、按钮失灵问题挑出来并解释清楚。适合写或调试任何 Streamlit app，尤其是遇到无限 rerun、刷新后值回到初始、或 callback/表单行为怪异时用。核心提醒你：每次交互都会从头跑完脚本，真正的状态只有 st.session_state 和当前控件值，缓存与 rerun、callbacks、forms、fragments 的交互都围绕这个模型展开。

▸ 展开 SKILL.md 英文原文

Streamlit execution model — reruns, session_state, callbacks, forms, caching, fragments. Use when writing or debugging ANY Streamlit app, especially infinite rerun loops, values that reset or vanish on refresh, buttons that "don't work", state errors, or live/auto-refreshing sections.

开发编程Streamlit执行模型状态管理通用

Stars

Forks

仓库内 Skill

7 日增星

安装 / 使用

给你的 Agent 一句话（通用）

帮我安装这个 skill：https://raw.githubusercontent.com/4k7yvjv7w4-ops/streamlit-agent-skills/main/streamlit-core/SKILL.md

或 curl 直取 SKILL.md

curl -fsSL "https://raw.githubusercontent.com/4k7yvjv7w4-ops/streamlit-agent-skills/main/streamlit-core/SKILL.md"

SKILL.MD 节选查看完整文件 ↗

# Streamlit core (1.58.x) — execution model, state, refresh

**The one mental model:** on EVERY user interaction (any widget touched),
Streamlit re-executes the WHOLE script top-to-bottom. There are no event
handlers, no persistent locals, no main loop. A run is a pure function of
`st.session_state` + current widget values. Everything below follows from this.

Runnable proof of every claim: `st_core_lab.py` in this skill folder
(`python -m streamlit run ~/.roo/skills/streamlit-core/st_core_lab.py`).
Verified on Streamlit **1.58**, API-checked on **1.55**: everything in
this skill exists on both (`st.fragment(run_every=)` needs ≥1.37).

## Hard rules (violations = the classic broken app)

1.

via SKILL·HOT · 数据来自 GitHub 公开信息 · 原文版权归作者所有