By rmzi
Launch multi-agent AI swarms for agentic SDLC workflows: decompose complex tasks into parallel subtasks, implement in isolated git worktrees, validate with full tests and linters, manage GitHub issues/PRs/reviews, fix bugs, bump semantic versions, and enforce code quality principles.
Codebase quality analyst. Use to scan for tech debt, code smells, missing tests, and file findings as GitHub issues.
Documentation specialist. Use when READMEs, changelogs, API docs, or inline documentation need updating after code changes.
Team lead for multi-agent tasks. Use when work needs decomposition, parallel execution, or coordination across agents and worktrees.
Deep codebase exploration. Use when you need thorough analysis of code, patterns, dependencies, or context before planning or implementation.
Code review specialist. Use to review diffs for quality, security, correctness, and best practices.
Fixing bugs using a test-first disciplined loop. Use when a bug is reported or discovered and needs a verified fix with minimal blast radius.
Bumping project version and updating changelog atomically. Use when releasing a new version — patch, minor, or major.
Save a checkpoint — bump version, commit, push. Use when work is complete and ready to ship without the full finish protocol.
Contributing changes to PDS skills, agents, or workflow. Use before modifying any PDS artifact to follow the whitepaper-alignment checklist.
Grounding in core development principles. Use when starting significant work, feeling stuck, or needing to reset decision-making clarity.
Matches all tools
Hooks run on every tool call, not just specific ones
Admin access level
Server config contains admin-level keywords
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.
Executes bash commands
Executes bash commands
Hook triggers when Bash tool is used
Hook triggers when Bash tool is used
Uses power tools
Uses Bash, Write, or Edit tools
Uses power tools
Uses Bash, Write, or Edit tools
A Claude Code plugin. Skills for consistency. Agents for scale. Install once, works across all projects.
PDS is deployed as a Claude Code plugin — skills, agents, hooks, and security settings. Install it to
~/.claude/plugins/pds/. Claude reads it, follows it, improves it.
Inside Claude Code:
/plugin marketplace add rmzi/portable-dev-system
/plugin install pds@pds-marketplace
Restart Claude Code. PDS skills and agents are now available across all projects.
curl -sfL https://raw.githubusercontent.com/rmzi/portable-dev-system/main/install.sh | bash
Installs the plugin to ~/.claude/plugins/pds/ and security settings to ~/.claude/settings.json.
Clean up old project-level files:
cd ~/your-project
curl -sfL https://raw.githubusercontent.com/rmzi/portable-dev-system/main/install.sh | bash -s -- --cleanup
See Migration Guide for details.
brew install jq / apt install jq)brew install python3 / apt install python3)git clone https://github.com/rmzi/portable-dev-system.git
cd portable-dev-system
make install # symlinks this checkout as the plugin
| Skill | Purpose |
|---|---|
/pds:swarm | Multi-agent team workflow (6-phase SDLC, lite/med/heavy tiers, branch merging) |
/pds:team | Agent roster, coordination, and dispatch modes |
/pds:grill | Requirement interrogation |
/pds:verify | Completion self-check |
/pds:finish | Branch completion protocol (includes quick ship) |
/pds:worktree | Git worktree workflow |
/pds:contribute | Contributing to PDS itself |
/pds:bugfix | Test-first bug fix loop |
/pds:bump | Version and changelog |
/pds:eval | Skill evaluation and testing |
/pds:rebase | Focused branch rebase |
/pds:pr-review | Address PR review comments |
/pds:pause | Save session state, WIP commit, resume later |
/pds:triage | Triage insights into GitHub issues across repos |
/pds:explore | Structural codebase queries via codebase-memory-mcp index (falls back to Grep) |
| Agent | Role | Model | Mode |
|---|---|---|---|
| orchestrator | Coordination — plans, decomposes, dispatches | opus | default |
| researcher | Deep codebase exploration | sonnet | plan |
| worker | Implementation in isolated worktrees | sonnet | acceptEdits |
| validator | Merge, test, verify acceptance criteria | sonnet | acceptEdits |
| reviewer | Code review — quality, security | sonnet | plan |
| documenter | Documentation updates | sonnet | acceptEdits |
| scout | PDS meta-improvements | haiku | acceptEdits |
| auditor | Codebase quality -> GitHub issues | sonnet | plan |
portable-dev-system/
├── .claude-plugin/plugin.json # Plugin manifest
├── agents/ # 8 agent definitions + shared-rules.md
├── skills/ # 15 skills (dir/SKILL.md format)
│ ├── swarm/SKILL.md
│ ├── team/SKILL.md
│ └── ...
├── hooks/hooks.json # Hook event handlers (see below)
├── hooks/scripts/ # Hook implementation scripts
├── scripts/ # Utility scripts (detect-patterns, efficiency-chart)
├── docs/ # Philosophy, whitepaper, team setup, source analysis
├── .claude/settings.json # Security settings (installed separately)
├── install.sh # Plugin installer
├── Makefile # make install
├── VERSION
└── CHANGELOG.md
PDS registers handlers for the following Claude Code hook events:
Hear your git context. Every repo has a voice. Every branch has a melody.
npx claudepluginhub rmzi/portable-dev-system --plugin pdsComplete 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.
AI-powered agents for specialized development tasks
Implementation planning, execution, and PR creation workflows with multi-agent collaboration
Language-agnostic development process harness implementing the Stateless Agent Methodology (SAM) 7-stage pipeline with ARL human touchpoint model and Voltron-style language plugin composition. Provides orchestration, workflows, planning, verification, and testing methodology that any language plugin can compose with.
Multi-agent orchestration framework for Claude Code, Gemini CLI, and Codex CLI — 19 agents, 14 skills, 16 commands, quality gates, TDD enforcement