anno-dsh-native
Run the following command in DeepSeek Harness:
dsh plugin install philmingdao/anno-dsh-native
Paste the following prompt into your AI chat to install this plugin:
Run dsh plugin install philmingdao/anno-dsh-native inside DeepSeek Harness to install this plugin; the full source is available at https://github.com/philmingdao/anno-dsh-native .
About this plugin
Reviewing and editing HTML files inside DeepSeek Harness traditionally means bridging an MCP server, spawning a separate npx child process, and launching an external HTTP service — a long chain with many moving parts. anno-dsh-native runs the entire Anno review workspace in-process within Harness, eliminating the MCP layer and external server so that HTML review, editing, and annotation become a single native call.
The plugin registers six html_review_* native tools and injects a full start, review, claim, deliver workflow into the system prompt. It launches a loopback-only HTTP server bound to 127.0.0.1, persists sessions to disk without ever overwriting the source file, and versions outputs incrementally. After a user completes text edits, format changes, element annotations, and page notes in the browser editor, clicking submit pushes the full review as a follow-up message straight back to the originating agent, which then claims the handoff and produces a verified standalone HTML file — no manual polling required.
If you build Agent workflows in DeepSeek Harness that involve HTML deliverables and want the review and editing step to stay local, dependency-free, and process-free, this is the lightest native integration available today.
Screenshots
Use Cases
- Review and annotate agent-generated HTML in a local browser editor, then push edits back to the agent
- Iterate on HTML drafts with versioned outputs, all running in-process with no external server
- Local-first HTML review workflow where every session and artifact is persisted to disk
Best For
- Developers building HTML-related Agent workflows in DeepSeek Harness
- Engineers who want local HTML review with zero external dependencies or child processes
- Teams that need seamless review handoff back to agents and automated delivery of final output
Related Plugins
ModLens is a vision plugin for DeepSeek Harness that gives text-only models sight by reading images pasted directly into chat, with zero-config setup and multiple vision engines.
On-demand vision for text-only DeepSeek Harness agents: built-in free keyless vision chain and 14 vision tools, routing image turns as tool calls to vision models with pixel fidelity, no Python needed, one-command install.
Give text-only models in DeepSeek Harness eyes, enabling image Q&A, long-screenshot OCR, UI restoration, and GUI visual tasks.
An armor-breaking plugin for DeepSeek: rewrites instruction priority, expands refusal-blocked phrases, persists across sessions, and shows a green active indicator.