Hypergraph is more powerful. Transform unstructured text into structured knowledge with LLMs. Graphs, hypergraphs, and spatio-temporal extractions — with one command.
chore: bump version to 0.10.2; news for 0.10.2
yifanfeng97479e74fd440a
diff --git a/README.md b/README.mdindex 9bd8987..fd6b242 100644--- a/README.md+++ b/README.md@@ -354,7 +354,7 @@ identifiers:## 📰 What's New-**v0.10.1** — 🔒 Pickle-free index storage: `AutoModel`/`AutoList` indexes are JSON now — legacy pickle indexes still load with a warning (`he build-index --force` migrates). *(#116)*+**v0.10.2** — 💬 Scoped chat (`he talk --source/--tag`) · 📖 method-selection guide + chunk vs graph comparison example · 🐛 observation-context persistence, MCP stdio crash & config-init fixes · 🏗️ GraphIndexMixin refactor.📰 **[Full release notes](https://yifanfeng97.github.io/Hyper-Extract/latest/news/)** · [All releases](https://github.com/yifanfeng97/hyper-extract/releases)diff --git a/README_ZH.md b/README_ZH.mdindex 8ff6e50..efacd1e 100644--- a/README_ZH.md+++ b/README_ZH.md@@ -354,7 +354,7 @@ identifiers:## 📰 最新动态-**v0.10.1** — 🔒 Pickle-free index storage: `AutoModel`/`AutoList` indexes are JSON now — legacy pickle indexes still load with a warning (`he build-index --force` migrates). *(#116)*+**v0.10.2** — 💬 范围对话(`he talk --source/--tag`)· 📖 方法选型指南 + chunk vs graph 对比示例 · 🐛 observation 持久化、MCP stdio 崩溃与 config init 修复 · 🏗️ GraphIndexMixin 重构。📰 **[完整版本说明](https://yifanfeng97.github.io/Hyper-Extract/latest/zh/news/)** · [全部 Releases](https://github.com/yifanfeng97/hyper-extract/releases)diff --git a/docs/en/news.md b/docs/en/news.mdindex 15c72e6..b7b8aff 100644--- a/docs/en/news.md+++ b/docs/en/news.md@@ -4,6 +4,16 @@ Release notes and highlights. For a complete changelog, see the [GitHub releases---+## v0.10.2 — Scoped Chat, Method Guide & Persistence Fixes++- **💬 Scoped chat** — `he talk --source/--tag` (and `ka.chat`) narrow the context to attributed/tagged documents, with capability detection for non-tracked types. *(#161, #160)*+- **📖 Choosing a method** — new "baseline vs structured" guide plus a side-by-side `chunk_rag` vs `graph_rag` example on the same corpus, with cost/latency comparison. *(#162, #163 — first two steps of #117)*+- **🏗️ Internal** — `GraphIndexMixin` extracts the shared merge/index/search surface of graph & hypergraph (−168 lines). *(#145)*+- **🐛 Fixes** — temporal/spatial observation context now survives dump/load (was silently lost); MCP load failures return as tool results instead of crashing the stdio session; `he config init` no longer writes a broken same-provider embedder for LLM-only providers; `Cog_RAG.chat` forwards `source_ids`/`tags`; `he list --lang zh` keeps method templates; document templates get output-shape validation; CLI banner lists jsonld/cypher. *(#167, #169, #165, #171, #159, #173, #155)*+- **📚 Docs** — resolved leftover conflict markers in MCP docs; unsmashed JSON-LD/Cypher sections; type-aware index-layout notes; Gemini provider guide. *(#149, #151, #153, #157)*++---+## v0.10.1 — Pickle-Free Index Storage- **🔒 Safe index format** — `AutoModel`/`AutoList` indexes are now stored as JSON (`index.json`: vectors + documents) and rebuilt in memory on load — **no pickle, no code execution**. Legacy pickle indexes (<= v0.10.0) still load with a deserialization warning; `he build-index --force` migrates them. *(#116)*diff --git a/docs/zh/news.md b/docs/zh/news.mdindex bff2878..53022d9 100644--- a/docs/zh/news.md+++ b/docs/zh/news.md@@ -4,6 +4,16 @@---+## v0.10.2 — 范围对话、方法选型指南与持久化修复++- **💬 范围对话** — `he talk --source/--tag`(及 `ka.chat`)可将上下文限定在已标注/打标签的文档内,对不支持的类型自动能力探测。*(#161, #160)*+- **📖 方法选型** — 新增"基线 vs 结构化"选型指南,以及同语料 `chunk_rag` vs `graph_rag` 的并排对比示例(含成本/延迟)。*(#162, #163 —— #117 的前两步)*+- **🏗️ 内部改进** — `GraphIndexMixin` 抽取 graph/hypergraph 共享的合并/索引/检索面(净删 168 行)。*(#145)*+- **🐛 修复** — 时空图谱 observation 上下文现在能跨 dump/load 保留(此前静默丢失);MCP 加载失败以工具结果返回,不再炸掉 stdio 会话;`he config init` 不再为纯 LLM provider 写入同源坏嵌入器;`Cog_RAG.chat` 透传 `source_ids`/`tags`;`he list --lang zh` 保留方法模板;document 模板增加输出结构校验;CLI banner 列出 jsonld/cypher。*(#167, #169, #165, #171, #159, #173, #155)*+- **📚 文档** — 解决 MCP 文档残留冲突标记;JSON-LD/Cypher 章节排版修复;索引布局说明按类型区分;Gemini provider 指南。*(#149, #151, #153, #157)*++---+## v0.10.1 — 索引存储脱离 pickle- **🔒 安全的索引格式** — `AutoModel`/`AutoList` 的索引改为 JSON 存储(`index.json`:向量 + 文档),加载时在内存中重建——**没有 pickle,不执行任何代码**。旧 pickle 索引(<= v0.10.0)仍可加载但会给出警告;`he build-index --force` 即可迁移。*(#116)*diff --git a/pyproject.toml b/pyproject.tomlindex 0d2d5fe..e0e1b25 100644--- a/pyproject.toml+++ b/pyproject.toml@@ -1,6 +1,6 @@[project]name = "hyperextract"-version = "0.10.1"+version = "0.10.2"description = "An intelligent, LLM-powered knowledge extraction and evolution framework with semantic search capabilities"readme = "README.md"requires-python = ">=3.11"diff --git a/uv.lock b/uv.lockindex 716a9f5..3baa0c0 100644--- a/uv.lock+++ b/uv.lock@@ -977,7 +977,7 @@ wheels = [[[package]]name = "hyperextract"-version = "0.10.0"+version = "0.10.2"source = { editable = "." }dependencies = [{ name = "faiss-cpu" },