dsh-docker
Run the following command in DeepSeek Harness:
dsh plugin install DevViking-Persike/dsh-docker
Paste the following prompt into your AI chat to install this plugin:
Run dsh plugin install DevViking-Persike/dsh-docker in DeepSeek Harness to install the plugin; the source lives at https://github.com/DevViking-Persike/dsh-docker .
About this plugin
When an agent needs to check container state, tail logs, or start and stop a Compose stack mid-conversation, the usual fallback is asking a human to open a terminal. dsh-docker removes that dependency by driving the Docker CLI already present on the machine. No daemon is launched at install time; reachability is probed on every call, so a host without Docker simply reports an unreachable engine instead of failing to load. The same plugin therefore works identically across dev laptops, CI runners, and test containers.
Five tools cover the core surface. The read-only trio docker_ps, docker_images, and docker_logs is active immediately after install and returns one structured line per container or image rather than raw CLI text. docker_logs keeps the newest characters when truncation kicks in and marks the cut so the model knows it is not seeing the full file. The two Compose lifecycle tools docker_compose_up and docker_compose_down change machine state, so they are opt-in: set compose to true in the config and they register. Arguments are passed as a fixed argv, never shell-interpreted, and docker_compose_down always removes the whole project to keep the blast radius predictable.
This plugin suits teams that have already built agent workflows in DeepSeek Harness and run Docker on the same machine. If your runbook includes checking container health, tailing a recent log to diagnose a failure, or bringing a Compose stack up and down during a debugging session, dsh-docker lets the agent close the loop inside the conversation without a round-trip to a human operator.
Use Cases
- An agent checks container status and tails recent logs mid-conversation
- An agent starts or stops a Compose stack while debugging
- An agent reads container output to pinpoint the root cause of a failure
Best For
- Teams running DeepSeek Harness workflows on Docker-equipped machines
- Workflow developers who need autonomous container diagnostics from an agent
- DevOps engineers who manage Compose stacks from a conversation context
Related Plugins
A method pack that makes AI coding agents plan against your real baseline, prove completion with fresh evidence, and reduce reworks and unsafe changes.
Turns the DeepSeek Harness session into a captain that builds a durable sub-agent team, splits goals into dependency-aware tasks, and coordinates work via direct messages and a live Web UI.
Gives coding agents design judgment, letting Claude Code, Cursor, and 70+ agents generate and iterate high-quality UI, presentations, and graphics right from the terminal.
Run the Pi ecosystem's plugins on DeepSeek Harness, unmodified, via a compatibility layer that implements Pi's public extension ABI on DSH's native services.