Orchestrates the full planner → architect → developer → reviewer pipeline for any development work — new features, bug fixes, refactors, or anything spanning more than one session.
Triggers the @reviewer agent on current uncommitted changes or a specified scope.
Summarizes work done since the last commit (or a given time window) in standup format.
Use when making structural decisions — new services, data models, API contracts, technology choices, or any change that affects more than one component. Invoke after planner, before developer. Also invoke after reviewer to validate findings and drive the fix loop until ARCH-APPROVED. Also invoke for ADR creation.
Use for all implementation work — writing code, fixing bugs, editing files, running commands. Always invoked after planner and/or architect on non-trivial work. For quick fixes and single-file changes, can be invoked directly.
Use when starting any non-trivial feature, task, or refactor. Breaks work into concrete, sequenced tasks with clear scope and acceptance criteria. Invoke before architect or developer on anything that spans more than one file or session.
Use after implementation is complete and before committing or merging. Reviews code for correctness, security, test coverage, and consistency with existing patterns. Also invoke on existing code when auditing for quality or security issues.
Template and conventions for documenting architectural decisions.
Format and criteria for capturing decisions, anomalies, and findings during development.
Branching, commit, and PR conventions for this project.
Coding patterns and conventions for this stack. Follow these exactly — consistency beats preference.
Requires secrets
Needs API keys or credentials to function
Uses power tools
Uses Bash, Write, or Edit tools
Own this plugin?
Verify ownership to unlock analytics, metadata editing, and a verified badge. GitHub access is read-only (username + org membership).
Sign in to claimOwn this plugin?
Verify ownership to unlock analytics, metadata editing, and a verified badge. GitHub access is read-only (username + org membership).
Sign in to claimBased on adoption, maintenance, documentation, and repository signals. Not a security audit or endorsement.
Personal Claude Code plugin monorepo. Each directory under plugins/ is a self-contained installable plugin.
| Plugin | Description |
|---|---|
dev-agents | Planner, architect, developer, reviewer + workflow commands + dev skills + GitHub MCP |
sre-toolkit | SRE agent + observability skills + deploy checklist + hooks |
docs | Shared team documentation — onboarding, conventions, runbooks |
1. Add this repo as a marketplace (once per machine):
/plugin marketplace add tiagokrebs/claude-agentic-platform
2. Install plugins:
# Every developer machine
/plugin install dev-agents@claude-agentic-platform
/plugin install docs@claude-agentic-platform
# Infra/SRE work only
/plugin install sre-toolkit@claude-agentic-platform
Requires GITHUB_TOKEN in your shell environment for the GitHub MCP server:
echo 'export GITHUB_TOKEN=ghp_yourtoken' >> ~/.zshrc
Decisions, anomalies, and findings captured automatically during development are pushed to:
github.com/tiagokrebs/claude-agentic-platform-engineering-intelligence
No setup required. The push-ei.sh script clones the repo to ~/claude-agentic-platform-engineering-intelligence on first run and pushes from there. Captures are written automatically by the architect and reviewer agents.
/plugin marketplace update claude-agentic-platform
claude-agentic-platform/
├── plugins/
│ ├── dev-agents/
│ │ ├── .claude-plugin/plugin.json
│ │ ├── .mcp.json ← GitHub MCP server
│ │ ├── CLAUDE.md ← agent routing rules
│ │ ├── agents/ ← planner, architect, developer, reviewer
│ │ ├── commands/ ← /plan, /review, /standup
│ │ ├── scripts/ ← session-summary.sh, push-ei.sh
│ │ └── skills/ ← git-workflow, adr-format, php-node-conventions, engineering-intelligence
│ ├── sre-toolkit/
│ │ ├── .claude-plugin/plugin.json
│ │ ├── agents/ ← sre
│ │ ├── commands/ ← /deploy-check
│ │ ├── hooks/ ← setup, pre-commit, post-write, notification
│ │ └── skills/ ← sre-patterns
│ └── docs/
│ ├── .claude-plugin/plugin.json
│ └── docs/
│ ├── onboarding.md
│ └── conventions.md
└── README.md
plugins/<name>/.claude-plugin/plugin.json with at minimum { "name": "<name>" }agents/, commands/, skills/, hooks/, .mcp.json as neededSRE agent, observability skills, deploy checklist, and hooks for aaPanel-managed infrastructure
npx claudepluginhub tiagokrebs/claude-agentic-platform --plugin dev-agentsComplete project development toolkit: 23 agents, 23 slash commands, 29 lifecycle hooks, and 69 reusable skills for Claude Code workflows
11 agents, 35 skills, 18 commands, 9 hooks — spec-driven multi-agent orchestration for Claude Code, with optional cross-device semantic memory.
Tool-agnostic agentic coding setup: 29 agents, 53 skills, 67 rules, 30 commands, 7 hooks, MCP servers, and a CLI-tool surface generated for 3 AI coding tools from a single canonical source. Counts derived from governance/inventory.json.
Make AI coding agents follow a repeatable engineering workflow with memory, verification, skills, and multi-agent setup
Portable Development System — AI-assisted development methodology with skills for consistency and agents for scale.
Tactical Agentic Coding course content by @IndyDevDan - Master agentic coding tactics to scale engineering and build self-operating codebases