AI Agent Hub
Back to plugins
⚙️

dsh-extras

Workflow Updated 2026.09.05

Run the following command in DeepSeek Harness:

dsh plugin install CatheadOwl/dsh-extras

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

Run dsh plugin install CatheadOwl/dsh-extras in your DeepSeek Harness terminal to install this plugin; the source repository is at https://github.com/CatheadOwl/dsh-extras .

About this plugin

Harness work in a dsh agent is, at its core, documentation work: the agent runs on a Markdown knowledge base, yet links rot quietly, project knowledge never makes it into the session context, and the base lacks a structured routing view. dsh-extras wraps dsh's turn-close hook into a composable quality-gate framework with a single registration face (registerGate) instead of every plugin grabbing the hook on its own, turning doc-quality maintenance from manual audits into an automated step that fires at the end of every turn.

Four independently toggleable modules each own a concern: gates orchestrates composable quality gates with a consecutive-block safety valve; markdown renames a file and rewrites every internal link in one move; prompt injects project knowledge into the session via declarative providers; and routes builds routing views over Markdown knowledge bases. Each module is an independent composition row in the host, so you can disable any one without affecting the rest.

If you drive a documentation-heavy workflow with dsh, maintain a non-trivial Markdown knowledge base, or want to add custom quality gates to the dsh plugin ecosystem, dsh-extras gives you a ready-to-use gate and knowledge-injection foundation. For plugin developers, it also exposes stable subpath APIs (such as gates/register) for extending the framework with your own gate definitions.

Use Cases

  • Automatically validate doc link integrity at the end of every turn
  • Inject critical project knowledge into the agent session context
  • Rewrite all internal links when renaming a Markdown file

Best For

  • Developers driving doc-heavy workflows with dsh agents
  • Teams maintaining large Markdown knowledge bases
  • Plugin authors extending the dsh quality-gate ecosystem