AI Agent Hub
Back to plugins
⚙️

dsh-ops-kit

Workflow Updated 2026.08.18

Run the following command in DeepSeek Harness:

dsh plugin install LeslieWylie/dsh-ops-kit

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

To install this plugin in DeepSeek Harness, use the command: dsh plugin install LeslieWylie/dsh-ops-kit, or visit the source repository at https://github.com/LeslieWylie/dsh-ops-kit for more installation options.

About this plugin

In automated agent workflows, agents can sometimes assert memory, plan completion, or benchmark results without evidence, leading to errors and unreliable behavior. The dsh-ops-kit plugin bundle addresses this by providing five read-only skills that force agents to show evidence before making key claims—for example, through memory searches that include source provenance or workflow plans that use staged verification. All tools are designed to be side-effect-free, performing only read operations to ensure safety and auditability.

Core capabilities include evidence-based orchestration planning, Git-first memory search, repository audit, and release hygiene checks, which together enhance transparency and reliability in agent processes. The bundle also integrates a runtime doctor tool to detect terminal prompt compatibility issues, preventing common failures. Its conservative design principle means it never silently creates issues or mutates repositories; all actions are explicit and reviewable.

This plugin is ideal for developers using the DeepSeek Harness ecosystem, teams engaged in multi-agent collaboration or benchmarking, and maintainers who need to ensure plugin release quality. Whether building reliable workflows, auditing repositories, or verifying release safety, dsh-ops-kit provides structured checkpoints to help users avoid the common trap of overconfident agents.

Use Cases

  • Verify the source and accuracy of agent memory
  • Plan multi-agent collaboration or benchmark workflows
  • Audit repository health before plugin releases

Best For

  • Engineers building multi-agent systems
  • Developers maintaining DeepSeek Harness plugins
  • DevOps teams performing rigorous verification