dsh-plugin-qiaomu-rss: Read RSS and Discuss It with AI Inside DeepSeek Harness
joeseesun/dsh-plugin-qiaomu-rss
Read and subscribe to RSS/Atom feeds inside DeepSeek Harness, switch between original, translated and rewritten articles, and discuss the current article in the host's native AI conversation.
dsh-plugin-qiaomu-rss gathers RSS subscriptions, article reading, and native AI conversation into one DeepSeek Harness interface, removing the back-and-forth between a feed reader and a chat window. You can search, filter, and add RSS/Atom sources on the explore page, import and export OPML, and switch between the original, translated, and rewritten article while keeping unread status, favorites, and search. Click the wand or select a passage to open the host's native conversation with the current article attached, turning your selection into an editable question draft. Maintained by joeseesun under GPL-3.0-only, it is currently verified on a local Desktop profile.
How to Install
dsh plugin --profile web add github:joeseesun/dsh-plugin-qiaomu-rss- Category
- Data & Knowledge
- Platform
- DSH Plugin
- Author
- joeseesun
- Distribution
- Plugin
dsh-plugin-qiaomu-rss Key Features
dsh-plugin-qiaomu-rss Repo Summary
What Does It Do?
Qiaomu RSS is a DSH plugin for DeepSeek Harness that brings curated and personal RSS/Atom feeds into the harness and pairs them with a native AI conversation. It removes the need to switch between a feed reader and a chat window when you want to ask about what you just read. Maintained by joeseesun under GPL-3.0-only, it is currently verified on a local Desktop profile and has not been published to npm.
Core Features
- Feed management: search, filter, and add RSS/Atom sources on the explore page, with OPML import and export and a bundled list of independent blogs.
- Multi-version reading: switch between the original, translated, and rewritten article, with unread tracking, favorites, search, and media content.
- AI companion reading: click the wand in the article toolbar or select text to open a native Harness conversation on the right, with the current article attached to the default workspace and the selection turned into an editable question draft.
- Resizable divider: drag the divider between article and chat, nudge it with arrow keys, or double-click to reset.
- Agent tools: eight rss_* tools covering channels, articles, search, subscriptions, refresh, and reading versions.
- Plugin settings: adjust reading appearance and the Qiaomu service address, manage subscriptions and OPML, and view about and donation information.
How to Use This Plugin?
After enabling it in DSH, add or import RSS/Atom sources on the explore page and open an article. To dig deeper, click the wand in the article toolbar or select a passage, and a native DeepSeek Harness conversation opens on the right with the current article attached; the selected text becomes an editable question draft. Reading and subscription state is stored in the Harness home under storages/qiaomu-rss/data.json, so upgrades do not wipe existing data. AI conversations use the host's configured model and account, with costs managed by the harness. Obsidian-style notes, highlight saving, and daily notes are intentionally absent.
How to Troubleshoot This Plugin?
The repository is verified on a local Desktop profile, and a clean-profile installation path has not yet been verified for a public release, so reproduce startup or update issues in a fresh profile first. After changing source files you must rebuild and commit the updated build output, otherwise the host keeps loading the old version; building and testing require Node.js 22 or higher. If subscription or reading data looks wrong, check the data file under the Harness home for integrity and repair it as the official documentation describes.
This page is an independent rewrite of the plugin's official README — for authoritative documentation and the latest changes, refer to the source: joeseesun/dsh-plugin-qiaomu-rss. The plugin is third-party code that runs on your machine once installed; inclusion does not imply endorsement — please review the source before installing.