dsh-usage-export
Run the following command in DeepSeek Harness:
dsh plugin install alenhu2005/dsh-usage-export
Paste the following prompt into your AI chat to install this plugin:
Run dsh plugin install alenhu2005/dsh-usage-export in your DeepSeek Harness terminal to install the plugin; the source repository is available at https://github.com/alenhu2005/dsh-usage-export
About this plugin
After a DSH turn, token usage, cost estimates, and inference stats tend to scatter across the interface, and the stats line is often clipped with an ellipsis. dsh-usage-export gathers all of that into one bar under the input box: per-turn and cumulative costs for uncached input, cached reads, and output tokens are listed separately; peak versus off-peak status is shown with the Taiwan-time peak windows (weekdays 09:00–12:00 and 14:00–18:00); and the full stat line—turns, steps, LLM time, time-to-first-token, throughput, cache hits—wraps naturally instead of being truncated.
For export, the plugin only reads visible on-screen text. It never touches reasoning traces, tool calls, attachments, or workspace files, and it sends no telemetry. You can export all messages, user-only, or assistant-only as Markdown, or invoke the browser-native print dialog to produce a PDF. A 5–20px slider scales headings, body text, and tables proportionally, making it easy to archive or hand off a clean document.
If you rely on DSH for multi-turn model inference, need to keep an eye on per-conversation costs in a familiar currency, or frequently export chat records for documentation, this local-first, MIT-licensed plugin fills the usage-and-export gap in the DSH Web UI.
Use Cases
- Track per-turn token costs and peak or off-peak status
- Export conversations selectively to Markdown or PDF for archiving
- View full untruncated inference metrics with wrapping
Best For
- Developers using DSH for multi-turn model inference in daily work
- Teams that need per-turn cost logs and inference metrics
- Users who frequently export conversations for documentation or sharing
Related Plugins
A unified suite combining hot runtime injection, task-aware thinking-mode routing, and a graded session protocol with red-team gates to sustain model diligence across long-horizon inference.
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.