AI Agent Hub
Back to plugins
⚙️

dsh-perfscope

Workflow Updated 2026.08.25

Run the following command in DeepSeek Harness:

dsh plugin install Evhye38496/dsh-perfscope

Paste the following prompt into your AI chat to install this plugin:

Install it in DeepSeek Harness by running dsh plugin install Evhye38496/dsh-perfscope; the source repository is available at https://github.com/Evhye38496/dsh-perfscope

About this plugin

The DeepSeek Harness plugin ecosystem is expanding at a blistering pace: skins, tools, and presets appear weekly. After you install a few dozen, the real questions surface. Which plugin is silently failing? Which one is eating tokens every session? Do you have duplicates, conflicts, or stale entries you forgot about? dsh-perfscope fills the gap that no other tool covers yet: a one-click, fully local health check that never sends your data off-machine.

It scores your setup from 0 to 100 across three weighted dimensions. Hard health catches mount failures, unmet dependencies, core version conflicts, and invalid configs. Runtime health tracks tool error and timeout rates, average latency against a baseline, and token or context share. Governance health flags overlapping capabilities, stale plugins, and sources outside curated registries. Every finding is phrased as a single actionable sentence and graded critical, warning, or suggestion. Fixes write a single patch row into cordis.patch.yml; they never uninstall, delete, or edit source files. Each write is previewed first, logged in the Doctor Change Log, and reversible with one click. Host-critical plugins and hand-edited rows are protected from accidental toggles.

It is built for anyone whose plugin stack has grown to a couple of dozen entries and started feeling fragile. CI and headless pipelines can run the same scan on schedule. And if you want to share a readable score card with teammates or the community, the report export is one click away. No network call, no cloud dependency: open it and you have an answer.

Use Cases

  • Diagnose which installed plugins are silently failing or slowing every session
  • Run periodic health scans in CI or headless pipelines with automated reporting
  • Export a readable 0-100 score card and actionable fix list to share with teammates

Best For

  • Heavy Harness users with large plugin stacks starting to feel fragile
  • CI and DevOps engineers who need headless environment audits on a schedule
  • Developers who want a quick, trustable verdict on their plugin tree