Run a multi-agent development workflow that plans, implements, reviews, documents, evaluates, and audits code changes through auto-discovered agents and commands, with quality gates and automated checks at each stage.
Save a manual checkpoint of the current workflow state with a summary.
View current workflow configuration settings.
Display a cost breakdown for the current or specified task.
**THIS COMMAND IS READ-ONLY (writes only to ~/.claude/doc-templates/).**
Run `python3 {__scripts_dir__}/crew-docs-import.py --dry-run` first and show the user what would be created. Ask for confirmation before running without `--dry-run`.
These standards apply to every crew agent. Each agent's own prompt file
> **Canonical data**: The WCAG checklist below has a machine-readable companion at [`skills/accessibility-reviewer/wcag-criteria.json`](../skills/accessibility-reviewer/wcag-criteria.json), including the contrast-ratio formula and threshold values. Claude Code hosts can invoke this as the `accessibility-reviewer` skill directly (see `skills/accessibility-reviewer/SKILL.md`). Other platforms read the prompt below verbatim. When the skill data and the prompt diverge, the JSON wins. Prefer running `axe-core` or `pa11y` for automated findings when available; this prompt covers what those tools can't (focus-order semantics, label clarity, cognitive load).
You are the **API Guardian**. Your job is to protect API contracts, detect breaking changes, ensure backward compatibility, and maintain API consistency.
You are a **Senior Architect** who reviews the Planner's implementation plan before it goes to implementation. You are the gate between planning and execution — the experienced voice that catches overengineering, missed opportunities, and flawed assumptions before they become expensive mistakes.
> **Canonical data**: The seven-category framework below has a machine-readable companion at [`skills/axiom-miner/heuristics.json`](../skills/axiom-miner/heuristics.json). Claude Code hosts can invoke this as the `axiom-miner` skill directly (see `skills/axiom-miner/SKILL.md`). Other platforms read the prompt below verbatim. When the skill data and the prompt diverge, the JSON wins.
WCAG 2.1 compliance review for UI components, forms, modals, navigation, and tables. Run on changes touching .tsx/.jsx/.vue/.html or component libraries. Catches accessibility barriers before they exclude users.
Surface hidden assumptions and unwritten rules before planning. Run on ambiguous or underspecified tasks where the planner would otherwise guess at conventions, data semantics, deployment context, or business rules.
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.
A multi-agent development workflow that orchestrates specialized AI agents through planning, implementation, and documentation phases. Supports custom crew definitions for any domain — software development, content creation, research, and more.
Supports Claude Code, GitHub Copilot CLI, Gemini CLI, OpenCode, Devin for Terminal, Droid (Factory.ai), and OpenAI Codex CLI.
New here? Check out the Visual Overview -- a jargon-free guide with diagrams, perfect for managers, team leads, and anyone curious about how it works.
Complex development tasks require multiple perspectives: architecture considerations, detailed planning, security review, edge case analysis, and careful implementation. Managing this manually with AI means constant context switching and oversight.
Agentic Workflow solves this by:
crew: section). Built-in software-dev crew is the default. See examples/crews//crew ask for second opinions without full workflow--mode)/crew workflows with Jira integration and custom post-setup commands| Platform | Install Guide |
|---|---|
| Claude Code | github.com/anthropics/claude-code |
| GitHub Copilot CLI | docs.github.com/copilot |
| Gemini CLI | github.com/google-gemini/gemini-cli |
| OpenCode | github.com/sst/opencode |
| Devin for Terminal | cli.devin.ai |
| Droid (Factory.ai) | factory.ai |
| OpenAI Codex CLI | github.com/openai/codex |
git clone https://github.com/Spiris-Innovation-Tech-Dev/agentic-workflow.git
cd agentic-workflow
./install.sh
Installs to:
~/.claude/commands/~/.claude/agents/~/.claude/scripts/ (helper scripts referenced by agents/commands)~/.claude/workflow-config.yamlgit clone https://github.com/Spiris-Innovation-Tech-Dev/agentic-workflow.git
cd agentic-workflow
.\install-copilot.ps1
Installs to:
.github/agents/ (repo-level) + ~/.copilot/agents/ (user-level)~/.copilot/workflow-config.yaml~/.copilot/mcp-config.json.github/copilot-instructions.mdgit clone https://github.com/Spiris-Innovation-Tech-Dev/agentic-workflow.git
cd agentic-workflow
./install-gemini.sh
npx claudepluginhub spiris-innovation-tech-dev/agentic-workflow --plugin agentic-workflowAgentic engineering done right — 57 structured workflows, 17 specialist agent personas, persistent memory across sessions, integrated learning partner, and impeccable UI design system. Works with Claude Code, Windsurf, Cursor, Gemini CLI, OpenCode, and Codex.
HelloAGENTS — The orchestration kernel that makes any AI CLI smarter. Adds intelligent routing, unified QA gates, safety guards, and notifications.
Repowire mesh usage skills for AI coding agents: cross-agent review and planning, delegate, usage patterns, and install/update. Backend-agnostic and parameterised on the agent you choose.
Agentic Delivery OS - AI-powered software delivery system
Plugin for effective agentic development
Autonomous session orchestrator for Claude Code - manages multi-phase development workflows