@zleap-ai/dsh-sag 0.1.0

package/CHANGELOG.md

@@ -0,0 +1,10 @@
+# Changelog
+
+## 0.1.0 - 2026-08-29
+
+- Validate the published bundle against the npm `latest` DeepSeek Harness `0.1.1-rc.2` release.
+- Publish dsh-sag as an installable DeepSeek Harness profile bundle.
+- Discover and configure a running local SAG through `setup` and `doctor`.
+- Expose search, reading, knowledge-source, upload, ingestion, and document-management tools.
+- Store connection metadata and credentials through dsh host services.
+- Keep the Python sidecar available as an explicit advanced embedded mode.

package/LICENSE

@@ -0,0 +1,21 @@
+MIT License
+
+Copyright (c) 2026 Zleap
+
+Permission is hereby granted, free of charge, to any person obtaining a copy
+of this software and associated documentation files (the "Software"), to deal
+in the Software without restriction, including without limitation the rights
+to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
+copies of the Software, and to permit persons to whom the Software is
+furnished to do so, subject to the following conditions:
+
+The above copyright notice and this permission notice shall be included in all
+copies or substantial portions of the Software.
+
+THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+SOFTWARE.

package/README.md

@@ -0,0 +1,58 @@
+# @zleap-ai/dsh-sag
+
+dsh-sag lets DeepSeek Harness use the local SAG personal knowledge base for search, reading, upload, note ingestion, and document management.
+
+Compatible with DeepSeek Harness `0.1.1-rc.2`, the current npm `latest` (there is no final `0.1.1` release yet). SAG must include the Connect dsh setting and local connector API. The unmodified SAG `1.8.3` release does not include this capability; use the build from [SAG PR #154](https://github.com/Zleap-AI/SAG/pull/154), or a later SAG release that includes it.
+
+## Quick start
+
+Install and start SAG first, then install the plugin:
+
+```sh
+dsh plugin --profile web add @zleap-ai/dsh-sag
+dsh plugin --profile web exec dsh-sag doctor
+```
+
+You can start using the plugin when `doctor` reports that SAG is connected. The command runs inside the `web` profile, so `dsh-sag` does not need to be on the system PATH. The plugin discovers common local addresses automatically; the normal setup does not require Python or manual credentials.
+
+## Connect SAG
+
+If `doctor` cannot find SAG, save a connection with any one of these routes:
+
+```sh
+# Discover a running local SAG automatically
+dsh plugin --profile web exec dsh-sag setup
+
+# Use a connection file exported by SAG
+dsh plugin --profile web exec dsh-sag setup ./sag-dsh.json
+
+# Use the local SAG address directly
+dsh plugin --profile web exec dsh-sag setup --url http://127.0.0.1:8000
+```
+
+Check the saved connection:
+
+```sh
+dsh plugin --profile web exec dsh-sag doctor
+```
+
+## Ask dsh naturally
+
+- “Search my SAG knowledge base for the DW-2412P30 upload limit and cite the source text.”
+- “Upload `/Users/me/Documents/product-manual.pdf` to SAG, then summarize it after processing finishes.”
+- “Save the following meeting decisions as a note in SAG: …”
+
+The plugin provides status checks, knowledge-source creation and listing, search and source-text reading, file upload, text ingestion, and document listing, reprocessing, and deletion. SAG advertises the operations available in the current installation; the plugin does not call capabilities that SAG has not enabled.
+
+## Update or remove
+
+```sh
+dsh plugin --profile web update @zleap-ai/dsh-sag
+dsh plugin --profile web remove @zleap-ai/dsh-sag
+```
+
+## Connection recovery
+
+Make sure SAG is running, then run `dsh plugin --profile web exec dsh-sag setup` and `dsh plugin --profile web exec dsh-sag doctor`. If automatic discovery fails, export `sag-dsh.json` from SAG or connect with the local-address command above.
+
+To let dsh manage the Python engine directly, see [advanced embedded mode](docs/embedded.md).

package/README.zh.md

@@ -0,0 +1,58 @@
+# @zleap-ai/dsh-sag
+
+dsh-sag 让 DeepSeek Harness 直接使用运行在本机的 SAG 个人知识库，包括检索、阅读、上传、写入和文档管理。
+
+适用于 DeepSeek Harness `0.1.1-rc.2`（当前 npm `latest`；目前没有正式的 `0.1.1` 稳定版）。SAG 需要包含“连接 dsh”设置和本地连接器 API；当前未修改的 SAG `1.8.3` 尚未包含该能力，请使用 [SAG PR #154](https://github.com/Zleap-AI/SAG/pull/154) 对应构建，或后续包含该能力的正式版本。
+
+## 开始使用
+
+请先安装并启动 SAG，再安装插件：
+
+```sh
+dsh plugin --profile web add @zleap-ai/dsh-sag
+dsh plugin --profile web exec dsh-sag doctor
+```
+
+`doctor` 显示“SAG 已连接”即可开始使用。命令由 `web` profile 执行，不需要把 `dsh-sag` 加入系统 PATH。插件会自动发现常见本机地址；不需要准备 Python 环境或手工填写密钥。
+
+## 连接 SAG
+
+如果 `doctor` 没有找到 SAG，任选一种方式保存连接：
+
+```sh
+# 自动发现已经启动的 SAG
+dsh plugin --profile web exec dsh-sag setup
+
+# 使用 SAG 导出的连接文件
+dsh plugin --profile web exec dsh-sag setup ./sag-dsh.json
+
+# 指定 SAG 的本机地址
+dsh plugin --profile web exec dsh-sag setup --url http://127.0.0.1:8000
+```
+
+配置完成后再次检查：
+
+```sh
+dsh plugin --profile web exec dsh-sag doctor
+```
+
+## 直接告诉 dsh
+
+- “在我的 SAG 知识库里查找 DW-2412P30 的上传限制，并给出原文依据。”
+- “把 `/Users/me/Documents/产品手册.pdf` 上传到 SAG，处理完成后总结主要内容。”
+- “把下面这段会议结论作为一篇笔记写入 SAG：……”
+
+插件提供状态检查、知识源创建与查询、知识检索与原文读取、文件上传、文本写入，以及文档查询、重处理和删除能力。SAG 会通过能力清单告诉插件当前可用的操作；未启用的能力不会被调用。
+
+## 更新或卸载
+
+```sh
+dsh plugin --profile web update @zleap-ai/dsh-sag
+dsh plugin --profile web remove @zleap-ai/dsh-sag
+```
+
+## 连接失败
+
+确认 SAG 正在运行，然后执行 `dsh plugin --profile web exec dsh-sag setup` 和 `dsh plugin --profile web exec dsh-sag doctor`。如果自动发现失败，可以从 SAG 导出 `sag-dsh.json`，或使用上面的本机地址方式连接。
+
+需要由 dsh 直接托管 Python 引擎时，请参阅[高级 embedded 模式](docs/embedded.md)。