dsh-cron
Run the following command in DeepSeek Harness:
dsh plugin install XiaoWind/dsh-cron
Paste the following prompt into your AI chat to install this plugin:
Run dsh plugin install XiaoWind/dsh-cron in your terminal, then restart the Web app to enable the /cron slash command; source is hosted at https://github.com/XiaoWind/dsh-cron .
About this plugin
Driving a recurring, scheduled loop in DeepSeek Harness used to mean re-prompting the agent by hand every time, which is easy to forget and hard to keep at a steady cadence. dsh-cron closes that gap with a single /cron slash command: type a standard five-field cron expression and the agent wakes at every wall-clock minute that matches, fires one tick toward your stated objective, and keeps going until you issue /cron stop.
The firing model is wall-clock anchored and non-interrupting. If a scheduled minute arrives while the agent is mid-turn, the tick simply waits for idle and catches up right after, so your current work is never broken. The schedule persists to a local JSON file, meaning a Web app restart does not lose it; the plugin resumes automatically on the next session. You can also append a natural-language objective after the expression (e.g. /cron "0 9 * * 1-5" fix the tests) so every periodic wake-up carries a clear direction, and tune the default cadence via config.defaultCron.
It is built for anyone who treats a DSH agent as a long-running on-call partner: periodic log triage, nightly test repair, recurring refactoring sprints, or any workflow that needs a gentle nudge at a fixed rhythm. One command to start, one to stop, no babysitting required.
Use Cases
- Wake the agent every 30 minutes to triage service logs
- Trigger test-fix tasks each weekday at 09:00
- Push refactoring forward every 10 minutes toward a stated goal
Best For
- Developers who want their agent on a fixed autonomous cadence
- Teams that need recurring automation without manual re-prompting
- Users treating DSH as a long-running workflow platform
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.