dsh-session-import
Run the following command in DeepSeek Harness:
dsh plugin install JuneLearn/dsh-session-import
Paste the following prompt into your AI chat to install this plugin:
Install this plugin in DeepSeek Harness by running dsh plugin install JuneLearn/dsh-session-import; the source code is available at https://github.com/JuneLearn/dsh-session-import.
About this plugin
You switch to another machine, another deployment, a fresh workspace — and the conversation you just had with DeepSeek Harness is stranded as a .zip or .jsonl export with no built-in way to bring it back. dsh-session-import closes that loop: it parses DSH /export archives, validates event structure and SHA-256 fingerprints, then registers the content as a brand-new DSH session, all through an Import conversation dialog on the new-session page with drag-and-drop and file-picker support.
Before you commit, a preview panel shows title, message counts, tool activity, turn structure, original workspace, model, and session state. You can pick a target workspace, rename the session, keep original timestamps or restamp to now, and selectively sync model and reasoning effort, agent preset, permissions, sandbox, approval policy, and plan mode. The plugin always allocates a fresh session ID and will never overwrite an existing one; if persistence, workspace attachment, or load validation fails at any point the entire import is rolled back so no partial sessions are left behind. On the ZIP side it enforces CRC-32, central-directory metadata, entry bounds, and local-header names, and outright rejects encrypted, multi-disk, and ZIP64 archives before anything touches the session store.
If you move between machines, share context across staging and production, or simply treat /export output as your session backup, this plugin turns that workflow into a reliable round-trip. It makes zero outbound network calls, never touches API keys, and parses uploaded files entirely in local memory. Combine dryRun=1 validation with open=1 instant resume and the migration friction drops to nearly nothing: verify first, then pick up exactly where you left off.
Use Cases
- Restore a prior conversation as a new session after switching machines or deployments
- Migrate context between environments while syncing model config and reasoning effort
- Rebuild a session in a fresh workspace from an /export backup and pick up where you left off
Best For
- DSH developers who switch between multiple machines or deployments
- Product and engineering teams migrating context across environments
- Heavy users who rely on /export for session backup and restore
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.