By sultano
General-purpose commands and agents for code review, git workflows, documentation, and continuous improvement
npx claudepluginhub sultano/claude-code-marketplace --plugin sultano-toolkitUse this agent when you have completed writing a logical chunk of code (a function, class, module, or feature) and want to ensure it follows software engineering best practices without being over-engineered. Examples: <example> Context: User has just implemented a new API endpoint handler. user: "I've just finished writing this API endpoint for user registration. Can you review it?" assistant: "Let me use the code-reviewer agent to analyze your code for best practices and potential over-engineering." [Uses Agent tool to launch code-reviewer] </example> <example> Context: User has refactored a class and wants feedback. user: "I refactored the UserService class to add dependency injection. Here's the code: [code snippet]" assistant: "I'll have the code-reviewer agent examine this refactoring to ensure it maintains good design principles without unnecessary complexity." [Uses Agent tool to launch code-reviewer] </example> <example> Context: User has completed a feature implementation. user: "Done implementing the payment processing feature. It has 3 new classes and 2 interfaces." assistant: "Great! Let me use the code-reviewer agent to review the implementation and check that it's well-designed without being over-engineered." [Uses Agent tool to launch code-reviewer] </example>
Use this agent to improve the project based on learnings from the current session. It analyzes what was learned, plans improvements, and implements changes that make future work more efficient. Delegates to project-scribe for documentation needs. Examples: <example> Context: After completing significant work. user: "Improve the project based on this session" assistant: "I'll launch the continuous-improver agent to apply learnings." <Task tool call to continuous-improver> </example> <example> Context: User wants to capture and apply session insights. user: "Extract learnings and make improvements" assistant: "I'll use the continuous-improver agent to improve the project." <Task tool call to continuous-improver> </example> <example> Context: End of development session. user: "What can we improve from this session?" assistant: "I'll launch the continuous-improver agent to identify and implement improvements." <Task tool call to continuous-improver> </example>
Use this agent for git operations, GitLab interactions (merge requests, issues, pipelines), and multi-repo management with meta. Handles commits, pushes, MRs, and cross-repo operations. Examples: <example> Context: User wants to commit and push changes. user: "Commit and push my changes" assistant: "I'll use the repo-manager agent to commit and push." <Task tool invocation to repo-manager agent> </example> <example> Context: User wants to create a merge request. user: "Create an MR for this branch" assistant: "I'll use the repo-manager agent to create a merge request." <Task tool invocation to repo-manager agent> </example> <example> Context: User wants to run a command across all repos. user: "Pull all repos" assistant: "I'll use the repo-manager agent to pull across all repositories." <Task tool invocation to repo-manager agent> </example>
Use this agent to manage project documentation. Responsibilities include writing/organizing docs in the docs/ folder, maintaining the index/barrel file, and ensuring docs stay synchronized with code. Note: For session learnings, use continuous-improver instead - it will delegate documentation tasks here when needed. Examples: <example> Context: User implemented a new feature. user: "I just added OAuth2 authentication" assistant: "I'll use the scribe agent to document this feature." <Task tool call to scribe> </example> <example> Context: User wants to reorganize documentation. user: "The docs folder is a mess" assistant: "I'll launch the scribe agent to restructure the documentation." <Task tool call to scribe> </example> <example> Context: Continuous-improver delegates documentation. user: [continuous-improver spawns scribe] assistant: "I'll document this process as requested by continuous-improver." <Task tool call to scribe> </example>
A curated collection of Claude Code plugins, commands, agents, and hooks by sultano.
Add the marketplace to Claude Code:
/plugin marketplace add sultano/claude-code-marketplace
Then install any plugin:
/plugin install <plugin-name>@sultano-plugins
Auto-approves safe commands and prompts for potentially dangerous ones. Reduces permission fatigue while keeping you in control.
/plugin install prompt-me-less@sultano-plugins
A general-purpose toolkit bundling commands, agents, and hooks for everyday development workflows.
/plugin install sultano-toolkit@sultano-plugins
/code-review -- Reviews all changed files against senior engineer standards. Checks for SOLID, DRY, KISS, and YAGNI adherence, plus bugs, type safety, performance, security, and test coverage. Provides actionable feedback with file paths and line numbers.
/commit-push -- Stages, commits, and pushes changes in one step. Analyzes your git log to match existing commit style, supports user-provided instructions, and never includes AI signatures.
code-reviewer -- An expert software architect agent that reviews completed code chunks (functions, classes, modules) for best practices. Specializes in catching over-engineering: unnecessary interfaces, factory patterns for single implementations, and abstraction layers that just pass data through.
continuous-improver -- Analyzes your session for repeated struggles, non-obvious dependencies, and configuration gotchas, then implements improvements directly. Adds BEHAVIOR comments to protect critical code, updates CLAUDE.md, and delegates larger documentation tasks to the scribe agent.
repo-manager -- Handles git operations, GitLab interactions (merge requests, issues, pipelines via glab), and multi-repo management (via npx meta). Enforces conventional commit format: type(scope): description.
scribe -- Sole owner of project documentation. Maintains docs/, README.md, and CLAUDE.md with cross-link integrity, content-title alignment, and contradiction-free consistency across all docs.
rtk-rewrite -- A PreToolUse hook that transparently rewrites CLI commands to use RTK for 60-90% token savings. Requires rtk >= 0.23.0 and jq.
statusline-powerline -- A powerline-style status line displaying billing mode, current folder, git branch with diff stats (+/-), model name, context window usage, and session timing. Uses a muted coral-to-purple gradient color scheme.
.claude-plugin/
marketplace.json # Marketplace registry
plugin.json # sultano-toolkit plugin metadata
commands/
code-review.md # /code-review slash command
commit-push.md # /commit-push slash command
agents/
code-reviewer.md # Code review agent
continuous-improver.md
repo-manager.md # Git/GitLab/meta operations agent
scribe.md # Documentation management agent
hooks/
rtk-rewrite.sh # RTK token-saving hook
statusline-powerline.sh
Comprehensive C4 architecture documentation workflow with bottom-up code analysis, component synthesis, container mapping, and context diagram generation
Uses power tools
Uses Bash, Write, or Edit tools
Core skills library for Claude Code: TDD, debugging, collaboration patterns, and proven techniques
Team-oriented workflow plugin with role agents, 27 specialist agents, ECC-inspired commands, layered rules, and hooks skeleton.
Upstash Context7 MCP server for up-to-date documentation lookup. Pull version-specific documentation and code examples directly from source repositories into your LLM context.
Complete collection of battle-tested Claude Code configs from an Anthropic hackathon winner - agents, skills, hooks, rules, and legacy command shims evolved over 10+ months of intensive daily use
Performance optimization suite with profiling, bundle analysis, and speed improvement tools