dsh-healthcheck

Heavy dsh plugin users can easily lose track of what is quietly degrading around them: the APFS data volume creeping past 90 percent, sessions and storages folders under ~/.dsh ballooning by tens of gigabytes, or network latency creeping into triple digits. By the time a hard failure forces you to debug, the cost far exceeds a routine check-up. dsh-healthcheck collapses all of that into a single prompt—say run a health check and it scans your environment entirely read-only, never touching or deleting a single file.

Each run covers seven categories: APFS data-volume utilisation and the byte-level delta since the last check, macOS memory pressure derived from vm_stat, outbound network latency, per-directory sizes under ~/.dsh (clipboard, sessions, profiles, storages), plugin versions compared against the npm registry, and availability of core toolchain binaries such as brew, node, python3, and git. Every finding is tagged CRITICAL, WARNING, or INFO, and results are automatically diffed against the previous snapshot—so a 6 GB overnight disk increase immediately surfaces as a yellow alert. History is persisted locally in a JSONL file (last 200 entries) with zero external calls.

It is built for developers who live inside the dsh plugin ecosystem, juggle multiple sessions and storage profiles, and want an early warning instead of a post-mortem. The plugin strictly reports; every remediation step remains an explicit, user-confirmed action.