dsh-routines
Run the following command in DeepSeek Harness:
dsh plugin install Jesse-njx/dsh-routines
Paste the following prompt into your AI chat to install this plugin:
Install with dsh plugin install Jesse-njx/dsh-routines in DeepSeek Harness; source code is available at https://github.com/Jesse-njx/dsh-routines.
About this plugin
Having an AI agent run your test suite at 2am, diagnose the worst failure, and leave a digest in your project sounds ideal, but manually spawning sessions, tracking output, and routing results where you need them is repetitive toil. dsh-routines collapses that into three steps: write a YAML file, attach a cron expression, and let the scheduler take over.
Each routine boots as an isolated one-shot Headless session with a full audit log you can replay later. The digest lands in the project by default and is pushed to WeChat or any other channel when a chat node is installed. Safety defaults are pragmatic and strict: unattended runs auto-deny any interactive approval, nested scheduling is disabled, and a hard timeout ensures a wedged agent cannot hold the repository past 9am. Delivery or summarizer failures are recorded on the run record, never thrown, so the scheduler itself never crashes on a single bad run. YAML files hot-reload on change, project-level routines override global ones by name, and five CLI commands cover the full lifecycle.
Built for developers and teams running DeepSeek Harness locally who want scheduled, auditable, cloud-free agent workflows.
Use Cases
- Nightly test-suite run with failure diagnosis and digest delivery
- Daily project status report pushed to a chat channel
- Weekly or monthly code-quality checks with results delivered as files
Best For
- Developers running DeepSeek Harness locally
- Small teams needing scheduled automation for agent tasks
- Engineers who prefer local deployment over cloud services
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.