dsh-maestro-review
Run the following command in DeepSeek Harness:
dsh plugin install ddtcorex/dsh-maestro-review
Paste the following prompt into your AI chat to install this plugin:
Run dsh plugin install ddtcorex/dsh-maestro-review in your terminal to install; the source is available at https://github.com/ddtcorex/dsh-maestro-review
About this plugin
Every team hits the same wall: merge requests pile up waiting for manual line-by-line review, standards drift between reviewers, and CI pipelines lack an automated quality gate. dsh-maestro-review plugs a complete automated MR review pipeline into DeepSeek Harness, covering GitLab webhook intake with deduplication and retention, spinning up reviewer and auditor child agents on disposable worktrees, and archiving structured findings, history, and signals per MR. The review model is overridable globally or per project, and while GitLab is the current ReviewProvider, the interface is designed so other forges can slot in behind the same contract.
Its headless CI mode deserves special attention: a shared reviewer project (holding secrets and the container image) can serve any number of source projects through a single lightweight bridge job, requiring no DSH host and no per-project mapping. Quick runs work diff-only; on-demand deep clones trigger a reviewer-only pass with static-only audit, and the container carries no govard runtime, keeping CI integration as thin as possible.
If you already run DeepSeek Harness across multiple projects and want consistent, auditable automated review coverage at the GitLab MR or CI stage, this piece of the Maestro suite closes the code-review gap. The settings UI injects directly into DSH Web slots, notification copy stays in this package while delivery routes through the optional maestroNotifier service, and configuration is shared through the namespaced settings store owned by dsh-maestro-config-lib. Plug it in and it works.
Use Cases
- Automated reviewer and auditor pipeline triggered on every GitLab MR submission
- Unattended code-quality gate inside a CI pipeline via a lightweight bridge job
- Unified review model configuration and findings archival across multiple projects
Best For
- Engineering teams running DeepSeek Harness across multiple projects who need automated review coverage
- Developers relying on GitLab MR workflows who want to reduce manual line-by-line review
- DevOps engineers seeking zero-host headless review via a shared reviewer image
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.