dsh-wemedia-workbench
Run the following command in DeepSeek Harness:
dsh plugin install Harzva/dsh-wemedia-workbench
Paste the following prompt into your AI chat to install this plugin:
Run dsh plugin install Harzva/dsh-wemedia-workbench in the DeepSeek Harness terminal; the source repository is https://github.com/Harzva/dsh-wemedia-workbench .
About this plugin
Managing self-media content spans asset discovery, line-by-line revision, and delivery to WeChat, Zhihu, Xiaohongshu, X, and more. These steps often scatter across tools and manual actions with no auditable thread. dsh-wemedia-workbench provides a local-first workbench for DeepSeek Harness that chains AI-assisted curation, immutable revisions, four-evidence reviews, native approvals, and persistent multi-channel jobs into a single closed loop: every action is traceable, approvals can halt the queue, and failures never silently rerun.
Key capabilities include read-only discovery of local articles, images, and videos with source mapping; revision-bound稿件 with diff previews and unsaved-change protection; four evidence categories (facts, editing, images & formulas, mobile visual) tied to specific revision bytes; batch WeChat draft dispatch capped at fifty items per run with per-article approval and queue-stop on rejection or uncertainty; adapter contracts for WeChat, Zhihu, Xiaohongshu, and X where the user supplies runtime bridges and credentials; plus local helper scripts for Matplotlib formula rendering and PyMuPDF chart cropping.
Built for independent creators and small content teams already working in the DeepSeek Harness environment who want to upgrade publishing from copy-paste to a workflow with approvals, evidence, and Job state. The project is at an early-development stage, locked to DSH 0.1.1-rc.2, is not an official DeepSeek product, and channel-bridge code does not guarantee verified posting on live accounts: validate before deploying.
Use Cases
- Curate local assets and papers into citation-preserved, immutable draft revisions
- Batch-deliver WeChat drafts with per-article native approval and queue-stop on rejection
- Prepare multi-channel jobs for Zhihu, Xiaohongshu, and X with preflight, execution, and target read-back
Best For
- Independent creators already working within the DeepSeek Harness environment
- Small content teams requiring auditable, approval-gated publishing workflows
- Writers upgrading self-media pipelines from copy-paste to tool-managed processes
Related Plugins
A method pack that makes AI coding agents plan against your real baseline, prove completion with fresh evidence, and reduce reworks and unsafe changes.
Turns the DeepSeek Harness session into a captain that builds a durable sub-agent team, splits goals into dependency-aware tasks, and coordinates work via direct messages and a live Web UI.
Gives coding agents design judgment, letting Claude Code, Cursor, and 70+ agents generate and iterate high-quality UI, presentations, and graphics right from the terminal.
Run the Pi ecosystem's plugins on DeepSeek Harness, unmodified, via a compatibility layer that implements Pi's public extension ABI on DSH's native services.