MIT-licensed, book-grounded AI code review skill pack for DeepSeek Harness — installs into dsh's ~/.dsh/skills as standard Agent Skills with six modes, cited findings and a 0–100 Health Score.
DSH integration
Compatible
Author-claimed
Safety audit
Unaudited
Last verified
2026-08-21
License
MIT
01What can it help you accomplish?
Run book-grounded AI code reviews inside DeepSeek Harness
Structured findings in Symptom → Source → Consequence → Remedy format with book citations, severity labels and a 0–100 Health Score
Developers and tech leads running dsh who want consistent, traceable code reviews instead of ad-hoc model opinions
Audit architecture and prioritize tech debt in a dsh session
Mermaid dependency graph with circular-dependency and Conway's Law checks, plus debt ranked by Pain × Spread with a Critical / Scheduled / Monitored repayment roadmap
Engineering teams onboarding a codebase, planning refactors, or deciding what debt to repay first
Sweep a repository and auto-fix decay risks from dsh
Unified R1–R6, T1–T6 and architecture scan with safe changes auto-applied, multi-file changes confirmed, architectural decisions flagged as manual, plus a Fix Log and score delta
Teams cleaning up a codebase before a release or migration
02How to install into DeepSeek Harness
Prerequisites
- DeepSeek Harness (`dsh`) installed — brooks-lint ships as Agent Skills that dsh loads from `~/.dsh/skills`
- bash + curl to run the one-line installer (or the repo's `scripts/install.sh`)
Installation steps
- 01
Run `curl -fsSL https://raw.githubusercontent.com/hyhmrright/brooks-lint/main/scripts/install.sh | bash -s -- dsh` (add `--project` to install into the current repo instead of your global config)
$ curl -fsSL https://raw.githubusercontent.com/hyhmrright/brooks-lint/main/scripts/install.sh | bash -s -- dsh
- 02
The installer copies the skills flat into `~/.dsh/skills` so the shared framework (`../_shared/`) always resolves
- 03
Just ask ("review this PR", "audit the architecture") to auto-trigger the matching skill from its description, or invoke the six commands from dsh's `/` menu — `/brooks-review` is auto-registered
Verify the integration
Not specified by the author
03DSH integration and capability boundaries
Installed into DeepSeek Harness (dsh) as standard Agent Skills via the one-line installer (`install.sh dsh`); dsh loads them from ~/.dsh/skills and auto-registers /brooks-review
Six review modes as slash commands / Agent Skills
a diff, changed files, or the whole repo, pointed at from a dsh session→mode-specific reports: PR review, architecture audit, tech debt, test quality, health dashboard, full sweep
/brooks-sweep applies fixes: safe changes are auto-applied to files, multi-file changes are confirmed, architectural decisions are flagged as manualCited structured diagnosis with Health Score
any production code or test suite, in any language→findings as Symptom → Source → Consequence → Remedy with book citations, severity labels (🔴/🟡/🟢) and a 0–100 Health Score
Project configuration via .brooks-lint.yaml
optional `.brooks-lint.yaml` in the project root→customized review behavior: strictness presets, disabled risk codes, severity overrides, ignore globs, custom risks and suppressions
reads `.brooks-lint.yaml` from the project root when present; all settings are optionalCI quality gates via GitHub Action
a pull request diff→review posted as a PR comment, optional fail-below / fail-on / fail-on-regression gates, SARIF output for GitHub Code Scanning
sends your diff to api.anthropic.com or the configured Anthropic-compatible endpoint (~$0.05–0.15 per PR run)posts PR comments, can upload SARIF to Code Scanning and fail the check
04Who is it for? When not to use it?
Good for
- Developers and tech leads running dsh who want consistent, traceable code reviews instead of ad-hoc model opinions
- Engineering teams onboarding a codebase, planning refactors, or deciding what debt to repay first
- Teams cleaning up a codebase before a release or migration
Not for
- The dsh integration is documented from dsh's official skill spec and verified at the file-layout level (the installer is tested), but has not yet been run end-to-end by the maintainer — only Claude Code, Gemini CLI and Codex CLI are maintainer-verified.
05Compatibility, maintenance and safety notes
- The dsh integration is documented from dsh's official skill spec and verified at the file-layout level (the installer is tested), but has not yet been run end-to-end by the maintainer — only Claude Code, Gemini CLI and Codex CLI are maintainer-verified.
- When used in CI, brooks-lint sends your diff to api.anthropic.com or whatever Anthropic-compatible endpoint you configure, so only point it at a host you trust with your source; each PR run costs ~$0.05–0.15.
- Review quality depends on the underlying model of the host agent — the README measures model quality separately with a 57-scenario eval suite, noting the figures vary slightly run to run.
MIT · actively maintained (latest release v1.5.0, 2026-08-14)
06Frequently asked questions
How do I install brooks-lint into DeepSeek Harness?
Run `curl -fsSL https://raw.githubusercontent.com/hyhmrright/brooks-lint/main/scripts/install.sh | bash -s -- dsh`. The installer copies the skills flat into `~/.dsh/skills`; add `--project` to install into the current repo instead of your global config. DeepSeek Harness also auto-registers `/brooks-review`.
Is it a native runtime or MCP?
Neither — brooks-lint ships as standard Agent Skills. dsh loads them from `~/.dsh/skills` (it also reads `~/.agents/skills` and `AGENTS.md`), and each skill auto-triggers from its description; for explicit invocation, dsh takes the command table as written, from its `/` menu or typed inline.
What prerequisites do I need?
Only DeepSeek Harness (dsh) itself — there are no other required dependencies. Optionally place a `.brooks-lint.yaml` in your project root to set strictness, disable risk codes or ignore files. The CI GitHub Action additionally needs an ANTHROPIC_API_KEY.
Where does my code go?
Inside dsh the review runs through the agent itself. In CI, the GitHub Action sends your diff to api.anthropic.com or any Anthropic-compatible endpoint you set via `api-base-url` — only point it at a host you trust with your source. Cost is ~$0.05–0.15 per PR run.
Has the dsh integration been verified?
Per the README, Claude Code, Gemini CLI and Codex CLI are maintainer-verified; dsh and eight other platforms are documented from each tool's official skill spec and verified at the file-layout level (the installer is tested), but not yet run end-to-end by the maintainer.
07Related DSH workflows
dsh-routing-suite
by yjh051108
dsh-routing-suite — injector + router-standard kit: install the runtime injector first, then the task-aware reasoning-mode router preset (measured P1-P23).
oh-dsh
by hust-open-atom-club
Community distribution: TUI, desktop, and Web UI as one bundle with layered installation.
perfect-web-clone
by ericshang98
Pixel-perfect clones of any webpage. Paste a URL, get a measured Vite + React replica.
dsh-super-injector
by yjh051108
BepInEx-style runtime mod injector: injects arbitrary local plugin packages into a running DSH web (junction + loader.create), with hot reload, staging, and uninstall tools.
08Data and sources
| DeepSeek Harness (`dsh`) | `~/.dsh/skills` | `~/.agents/skills`, `AGENTS.md` | [setup](docs/dsh-setup.md) |
Kiro, Factory Droid, and DeepSeek Harness also auto-register `/brooks-review`.
# <platform> = opencode · cursor · windsurf · antigravity · pi · kiro · copilot · droid · dsh · gemini · codex · agent…
This page is generated from the project’s public documentation, repository metadata and a structured parse of DSH Plugins; last verified on 2026-08-21. Found an error? Submit a correction.
Best DeepSeek Harness Plugins
Twelve plugins worth installing first — picked from the whole catalog, across every category.