dsh-loop
Run the following command in DeepSeek Harness:
dsh plugin install XiaoWind/dsh-loop
Paste the following prompt into your AI chat to install this plugin:
In DeepSeek Harness, run dsh plugin install XiaoWind/dsh-loop to install the plugin; the source code is hosted at https://github.com/XiaoWind/dsh-loop.
About this plugin
When an Agent must repeat the same kind of task over and over, triggering each cycle by hand is both tedious and easy to miss. dsh-loop replaces that friction with a single /loop slash command that puts the Agent on a fixed-interval autonomous loop, so you can step away from the console.
The core experience is simple: /loop accepts a flexible duration token built from ms, s, m, h, and d units (for example 90s or 1h30m), optionally followed by a free-text objective that anchors every tick. The very first tick fires immediately; each subsequent tick is scheduled only after the Agent has returned to idle, guaranteeing that a running turn is never interrupted. Loop state is persisted to disk, so a session restart does not lose progress — /loop resume restores the saved loop in one step, and /loop stop discards it cleanly.
It is built for developers whose workflows call for long-running, periodic Agent work: continuously fixing flaky tests, polling external services for status changes, or iteratively refining a large refactoring. One command keeps the Agent working steadily in the background while you focus on the outcome rather than the cadence.
Use Cases
- Poll external services on a fixed cadence and trigger follow-up actions whenever state changes
- Run a repetitive test-fixing cycle until every case passes, stopping manually when done
- Iterate on a large refactoring over an extended window, each tick focused on the stated objective
Best For
- Developers who need an Agent to run periodic, self-scheduled tasks in the background
- Users who prefer one command over repeatedly retyping prompts to drive the same action
- Engineers orchestrating long-running Agent workflows through the DSH Web interface
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.