bug-ops/claude-plugins

Letzte erfasste Quellaktivität: 8. Oktober 2026 um 16:31

SkillsMP-Katalog aktualisiert: 5. November 2026 um 19:57

gesammelte Skills: 20

GitHub-Stars: 11
GitHub-Forks: 0

Mit einem KI-Assistenten installieren

Diesen Prompt kopieren und in den verwendeten KI-Assistenten einfügen.

Follow https://skillsmp.com/skill-install/prompt.md to install Agent Skills from https://github.com/bug-ops/claude-plugins.

README des Repositorys

README-Quelldatei: README.md

README-Quelldatei lesen

Vollständige README anzeigen

# Claude Plugins Collection A curated collection of specialized plugins for Claude Code CLI, designed to enhance development workflows with domain-specific expertise. ## Overview This repository contains plugins that extend Claude Code's capabilities with specialized agents, tools, and workflows for various development domains. Each plugin provides a set of expert agents tailored to specific aspects of software development. ## Available plugins ### dev-agents (`dev-agents`) [![Version](https://img.shields.io/badge/version-2.2.0-blue)](./dev-agents) [![License](https://img.shields.io/badge/license-MIT-green)](./dev-agents/LICENSE) Language-agnostic development agents covering the whole development lifecycle, with stack profiles for **Rust**, **TypeScript/JavaScript**, and **Luau/Lua** (including Roblox). Agents carry the process; the `stack` skill detects the project's language and loads the matching rules: toolchain commands, idioms, anti-patterns, and checklists per role. Polyglot repositories load every matching profile. Formerly `rust-agents` (1.x). See the [migration notes](./dev-agents/README.md#migrating-from-rust-agents-1x). **Location**: [`./dev-agents`](./dev-agents) **Key features**: - 15 specialized agents (opus/sonnet) that apply per-language rules from the `stack` skill - 20 productivity skills: - **stack** — Stack detection and per-language reference loader (Rust, TypeScript, Luau/Lua); loaded by every agent at startup - **team-develop** — Multi-agent development orchestration with peer-to-peer communication; Step 0 classifies tasks into one of nine chains (`new-feature`, `spec-driven`, `bug-fix`, `refactoring`, `security`, `docs`, `dependency`, `performance`, `ci-cd`) and runs only the agents that chain needs - **team-debug** — Multi-agent root cause investigation: debugger → parallel review → consolidated report → user decides next steps - **agent-handoff** — Inter-agent context sharing - **solve-issue** — Solve GitHub issues end-to-end via worktree + team-develop - **triage-and-solve** — Triage open issues by priority, group, and solve - **continuous-improvement** — Orchestrates live-tester + researcher + arch-analyst + security-analyst agents for a full CI cycle - **live-testing** — Live execution of the built artifact, anomaly detection, coverage tracking, bug filing - **research-protocol** — Dependency monitoring, research & innovation, competitive parity - **arch-inspect** — Architecture and code-quality audit protocol (type safety, modularity, testability, readability, DRY, async) - **security-audit** — Vulnerability audit protocol (dependency advisories, unsafe and escape-hatch code, secrets, injection, crypto, auth, crash-DoS, supply chain) - **init-project** — Scaffold project infrastructure for the plugin (Rust, TypeScript, Luau/Lua, polyglot) - **release** — Release preparation: version bump, changelog, docs refresh - **readme-generator** — Professional README generation - **mdbook-tech-writer** — Technical documentation with mdBook - **obsidian-zettelkasten** — Obsidian knowledge base with Zettelkasten method - **sdd** — Spec-Driven Development workflow - **spec-from-stream** — Business requirements from stream-of-consciousness (BRD/SRS/NFR) - **fast-yaml** — YAML validation, formatting, and conversion - **rust-modern-apis** — Lookup table for stable Rust APIs added in 1.89–1.99 (loaded by the Rust profile) - rust-analyzer, typescript-language-server, luau-lsp, and lua-language-server LSP integration - Proactive triggers for automatic agent selection **Agents included**: | Agent | Model | Specialization | |-------|-------|---------------| | architect | opus | Type-driven architecture, module and package boundaries, strategic decisions | | developer | sonnet | Idiomatic, type-safe feature implementation | | testing-engineer | sonnet | Test coverage, test infrastructure, benchmarks | | performance-engineer | sonnet | Profiling, optimization, build speed | | security-maintenance | opus | Security scanning, vulnerability assessment, dependency management | | code-reviewer | sonnet | Quality assurance, standards compliance, code review | | cicd-devops | sonnet | GitHub Actions, cross-platform testing, workflows | | debugger | sonnet | Error diagnosis, runtime debugging, crash analysis | | critic | opus | Adversarial design critique, assumption stress-testing | | sdd | sonnet | Full-cycle SDD orchestrator: BRD/SRS/NFR → spec/plan/tasks → knowledge base | | live-tester | sonnet | Live execution, anomaly detection, coverage tracking, bug filing | | tech-writer | sonnet | User-facing documentation with mdBook, progressive disclosure | | researcher | sonnet | Dependency monitoring, security advisories, research, competitive parity | | arch-analyst | opus | Architecture audits: type system anti-patterns, DRY violations, structure, async defects | | security-analyst | opus | Vulnerability audits: dependency advisories, unsafe code, secrets, injection, crypto misuse, auth, supply chain | **Best for**: Rust, TypeScript, and Luau/Lua projects, Roblox experiences and polyglot repositories included, requiring expert guidance in architecture, performance, security, testing, DevOps, or multi-agent team workflows. [→ Read full documentation](./dev-agents/README.md) ## Installation ### Quick start: Install from marketplace The easiest way to install plugins is via the marketplace: ```bash # Add the marketplace claude plugin marketplace add bug-ops/claude-plugins # Install the dev-agents plugin claude plugin install dev-agents@claude-dev-agents ``` This method provides automatic updates and centralized plugin management. ### Alternative: Install from local directory For development or testing, load the plugin directly from a local path: ```bash # Load for one session claude --plugin-dir ./dev-agents # Or install from the local directory claude plugin install /path/to/claude-plugins/dev-agents ``` ### Prerequisites - [Claude Code CLI](https://docs.claude.com/claude-code) installed and configured - The toolchain of the stack you work in: - Rust: Rust 1.85+, rust-analyzer for LSP support - TypeScript: Node.js LTS, the project's package manager, typescript-language-server for LSP support - Luau/Roblox: the tools pinned in the project's `rokit.toml` (Rojo, Wally or pesde, StyLua, selene, luau-lsp, Lune); Lua: the target Lua version, LuaRocks, lua-language-server ## Usage Once installed, agents from the plugins become available in Claude Code: ```bash # Start Claude Code claude # View available agents /agents # Agents will be automatically suggested based on your task ``` ### Example workflow ``` User: "I want to create a new Rust web service with database integration" Claude: → /team-develop detects stack rust, classifies as new-feature → architect → critic → developer → validators → reviewer → commit User: "Add rate limiting to the Express API" Claude: → /team-develop detects stack typescript, classifies as new-feature → same chain; every agent loads the TypeScript profile User: "Add a trading system between players to my Roblox game" Claude: → /team-develop detects stack luau (Roblox mode), classifies as new-feature → same chain; remote validation, session-locked persistence, and server authority come from the Luau profile User: "Draft a spec for a multi-tenant billing module — no code yet" Claude: → /team-develop classifies as spec-driven → architect → critic → sdd → reviewer → commit spec → open follow-up implementation issue User: "My service is timing out under load" Claude: → /team-debug: debugger investigates → parallel review (arch, critic, security, perf) → consolidated report (root cause + critical fixes + follow-up items) → user decides next steps ``` > [!TIP] > Agents can delegate work to other agents using the handoff protocol, preserving context between transitions. ## Repository structure ``` claude-plugins/ ├── README.md # This file ├── .gitignore ├── .claude-plugin/ │ └── marketplace.json # Marketplace catalog ├── .local/ # Working documents and reports (gitignored) ├── dev-agents/ # dev-agents plugin (includes team orchestration) │ ├── README.md │ ├── .claude-plugin/ │ │ └── plugin.json │ ├── .lsp.json # rust-analyzer, typescript-language-server, luau-lsp, lua-language-server │ ├── .devcontainer/ │ ├── agents/ # 15 specialist agents │ └── skills/ # 20 skills; stack/references/{rust,typescript,luau}/ hold the language rules └── [future-plugins]/ # Additional plugins ``` ## Marketplace This repository provides a Claude Code plugin marketplace at `.claude-plugin/marketplace.json`. ### Using the marketplace ```bash # Add marketplace from GitHub claude plugin marketplace add bug-ops/claude-plugins # List available plugins claude plugin list # Install a plugin claude plugin install dev-agents@claude-dev-agents # Update marketplace and plugins claude plugin marketplace update claude-dev-agents ``` ### Adding to project settings For team use, add the marketplace to `.claude/settings.json`: ```json { "extraKnownMarketplaces": { "claude-dev-agents": { "source": { "source": "github", "repo": "bug-ops/claude-plugins" } } }, "enabledPlugins": { "dev-agents@claude-dev-agents": true } } ``` This ensures team members are prompted to install the marketplace and plugins when they trust the project. ## Development environment ### Using DevContainer Each plugin may include DevContainer configurations for isolated development. See individual plugin documentation for details. ## Plugin development guidelines When creating new plugins for this repository: 1. **Structure**: - Each plugin in its own directory - Include `.claude-plugin/` configuration - Provide comprehensive README.md - Use `.devcontainer/` for development environment (optional but recommended) 2. **Documentation**: - All documentation in English - Clear usage examples - Installation instructions - Requirements and dependencies 3. **Agents**: - Focused, single-responsibility agents - Clear specialization boundaries - Appropriate model selection - Distinct color coding for easy identification - Include handoff protocol for multi-agent workflows 4. **Best practices**: - Follow [Microsoft Rust Guidelines](https://microsoft.github.io/rust-guidelines/agents/all.txt) for Rust-related plugins - Use `.local/` directory for working documents - Include version information - Add comprehensive examples ## Contributing Contributions are welcome! To add a new plugin or improve existing ones: 1. Create a new directory for your plugin 2. Include `.claude-plugin/` configuration 3. Write comprehensive documentation 4. Test thoroughly with Claude Code 5. Submit a pull request To add a language to dev-agents, add `dev-agents/skills/stack/references/<language>/` with a `toolchain.md` and one file per role, then extend the detection table in `stack/SKILL.md`. ## Roadmap Future ideas: - More stack profiles for dev-agents: Python, Go - Database management and optimization - Cloud infrastructure (AWS, Azure, GCP) ## License MIT

README des Repositorys wird geladen…

Skills in diesem Repository

Klassifizierung ausstehend

nicht klassifiziert

Triage open GitHub issues by priority (P0-P4 labels first, category labels as fallback), group compatible ones into a single PR, then solve via /solve-issue.

Quellsprache: Englisch

2026-10-08

Es werden 20 von 20 gesammelten Skills angezeigt.