dsh-request-privacy
Run the following command in DeepSeek Harness:
dsh plugin install jhckevin/dsh-request-privacy
Paste the following prompt into your AI chat to install this plugin:
Run dsh plugin install jhckevin/dsh-request-privacy in the terminal where DSH is hosted; the open-source repository is at https://github.com/jhckevin/dsh-request-privacy
About this plugin
Every chat request sent through DeepSeek Harness carries extra correlated fields in the header — the Harness user ID, session identifier, and context-compaction markers. These are not required for the chat to function, yet they create additional traceable linkage beyond what the conversation itself already transmits.
dsh-request-privacy adds a request-privacy toggle to the DSH settings page. When enabled, the next DeepSeek request switches to a private-client identity and drops the associated fields; when disabled, native headers are restored and everything proceeds as before. The toggle takes effect on the very next request, applies to existing conversations, and requires no restart.
It is aimed at users who are sensitive to request-level metadata and want to minimise the correlated information sent to the service provider. Note that it is not an official do-not-use-for-training switch: API keys, chat content, and tool descriptions are still transmitted, and the plugin does not cover file uploads, telemetry, or third-party providers.
Screenshots
Use Cases
- Chat requests no longer carry Harness user ID, session ID, or compaction markers
- Toggle compact headers mid-session without recreating any conversation
- Disabling the switch restores native headers on the very next request with no disruption
Best For
- DeepSeek Harness users who prioritise request-metadata minimisation
- Developers seeking fewer correlated fields in outgoing requests
- Community members comfortable with the DSH plugin ecosystem
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.