Help us improve
Share bugs, ideas, or general feedback.
Share bugs, ideas, or general feedback.
Share bugs, ideas, or general feedback.
By stevenmays
Enforce extreme code-quality standards across a repo—audit maintainability, distill PR review feedback into automated pre-review checks, block destructive Git commands and secrets in CI hooks, and generate images from prompts via Gemini.
npx claudepluginhub stevenmays/dotfiles --plugin maysDistill coding standards from the repo's recent merged PRs into .claude/standards.md
Extremely strict maintainability audit of the current branch (code-judo, 1k-line rule, spaghetti growth)
Resolve all merge conflicts non-interactively, then build and test
Distill human review feedback on your PR into the Manual section of .claude/standards.md
Review the current branch against distilled repo standards before opening a PR
Run an extremely strict maintainability review for abstraction quality, giant files, and spaghetti-condition growth. Use for an extreme code quality review, deep code quality audit, or especially harsh maintainability review.
Generate images using Google Gemini with customizable options
Patterns for AWS Lambda, DynamoDB, SQS, and Secrets Manager. Use when working on serverless AWS projects.
Write in Steven's voice—pragmatic, curious, pedagogical. Opens with measurable payoffs, builds mental models from first principles, uses worked examples, and handles uncertainty honestly. Use for essays, blog posts, and technical articles.
Executes bash commands
Hook triggers when Bash tool is used
Modifies files
Hook triggers on file write and edit operations
Uses power tools
Share bugs, ideas, or general feedback.
Own this plugin?
Verify ownership to unlock analytics, metadata editing, and a verified badge.
Sign in to claimOwn this plugin?
Verify ownership to unlock analytics, metadata editing, and a verified badge.
Sign in to claimBased on adoption, maintenance, documentation, and repository signals. Not a security audit or endorsement.
Review pull requests with structured analysis and approve with confidence
Review PR or staged changes via specialist agents. Synthesizes a unified severity-ranked report.
PR review, interactive git diff annotation review, and writing style guide
Automated code review for pull requests using multiple specialized agents with confidence-based scoring
The Paranoid Android — quality tools for code review, knowledge compounding, and work execution
Comprehensive PR review agents specializing in comments, tests, error handling, type design, code quality, and code simplification
Uses Bash, Write, or Edit tools
Uses Bash, Write, or Edit tools
Share bugs, ideas, or general feedback.
Personal Claude Code configuration, packaged as an installable plugin (mays).
From within Claude Code:
/plugin marketplace add stevenmays/dotfiles
/plugin install mays@dotfiles
Then sync personal settings (optional, overwrites ~/.claude/settings.json):
./sync.sh
After new commits land in this repo, refresh the marketplace catalog — Claude Code re-pulls the source and upgrades the installed mays plugin in the same step:
/plugin marketplace update dotfiles
Then run /reload-plugins when prompted (or restart Claude Code) to load the new version. To make this automatic, enable auto-update for the marketplace under /plugin → Marketplaces → dotfiles.
Non-interactive equivalent:
claude plugin marketplace update dotfiles
| Command | Purpose |
|---|---|
/distill-standards [N] | Mine the last N merged PRs (asks how many, default 50) for review patterns, distill into .claude/standards.md |
/extreme-code-quality-review | Extremely strict maintainability audit of the current branch, standards-aware |
/fix-merge-conflict | Resolve merge conflicts non-interactively |
/learn-from-review [PR#] | Distill human review feedback on your PR into .claude/standards.md |
/pre-review | Check the branch diff against distilled standards (with staleness warning), then offer to apply fixes |
/security-audit | Audit home-directory dotfiles for security issues (read-only report) |
/test-and-fix | Run tests and fix failures until green |
| Skill | Purpose |
|---|---|
extreme-code-quality-review | Rubric for the strict maintainability audit (code-judo, 1k-line rule, spaghetti) |
gemini-image-generator | Generate images via Gemini API |
serverless-aws | AWS Lambda/DynamoDB/SQS patterns |
writing-style | Personal writing voice for technical content |
| Agent | Purpose |
|---|---|
extreme-code-quality-review | Runs the strict audit in a subagent to protect the main context |
| Hook | Event | Purpose |
|---|---|---|
block-destructive-git | PreToolUse (Bash) | Block force pushes, hard resets, and other destructive git ops |
block-hardcoded-secrets | PreToolUse (Edit/Write) | Block edits that introduce hardcoded API keys/passwords |
post-edit-format | PostToolUse (Edit/Write) | Auto-format edited files with Prettier when the project uses it |
/distill-standards # once per repo (asks how many PRs, default 50)
/pre-review # before every PR (warns when standards are stale)
/learn-from-review <PR#> # after a human reviews your PR
distill-standards reads the repo's merged PR history via gh, extracts recurring review feedback and conventions (distilled rules, not raw comments), and writes .claude/standards.md. pre-review checks your branch against those rules plus a baseline AI-slop checklist (obvious comments, gratuitous defensive checks, as any, single-use abstractions), then offers to apply the fixes. learn-from-review closes the loop: it distills the human feedback on your PR into the ## Manual section of standards.md, which distill-standards carries forward on regeneration — so pre-review catches that feedback before the next reviewer has to.
Every review surface consults .claude/standards.md when it exists: /pre-review and /extreme-code-quality-review load it directly, and sync.sh maintains a managed block in ~/.claude/CLAUDE.md so the native /code-review picks it up in any repo.
/pre-review # standards conformance — catches what reviewers flagged before
/code-review high # native correctness review — bugs, not style
/extreme-code-quality-review # structural audit — only for large or structural changes
Then commit, push, and open the PR (plain request — conventions live in the ~/.claude/CLAUDE.md managed block). After human review lands, /learn-from-review.
settings.json (synced to ~/.claude/settings.json by sync.sh; the script also maintains a managed block in ~/.claude/CLAUDE.md with the standards lookup and git conventions):
includeCoAuthoredBy: false — no co-author tags in commitsalwaysThinkingEnabled: truepermissions.allow — pre-approved safe commands (git status/diff/log, npm run/test, prettier, eslint, tsc, gh)enabledPlugins — marketplace plugins, including this one