dsh-double

An agent edits its own config, restarts, and immediately breaks. That is the moment every DeepSeek Harness user dreads. dsh-double closes the loop: it applies a change set to a standby copy first, runs static checks and a trial boot, and only then flips to the main instance. If the main fails to come back healthy, it auto-rolls back to the last-known-good snapshot. You stop needing an outside agent to rescue the one that broke itself.

Beyond blue-green self-evolution, the plugin adds process-level health probing across TCP, HTTP, and process checks, an on-demand cold standby that wakes as a spare brain when mechanical recovery fails, and a fully isolated second dsh instance called the testbed. Each side runs its own LLM agent, and the two converse over the official session API. The testbed proves an evolution in its sandbox first; once the change is validated, the twin agent blue-greens it back onto the main. Every exchange lands in a shared, replayable transcript, so the settings dashboard shows the full two-way conversation at a glance.

If you run long-lived agent workloads on dsh and want your agent to iterate on its own config or workflow without dropping context, dsh-double turns self-evolution from a risky stunt into a reliable daily routine.

It is built for power users who deploy dsh in production-like environments, run multi-agent pipelines, and need an auditable trail of every self-modification. The zero-dependency CLI and out-of-process supervisor mean the life-support layer survives even when dsh itself cannot boot, giving you a safety net that no in-process plugin alone can provide.