Help us improve
Share bugs, ideas, or general feedback.
Share bugs, ideas, or general feedback.
Share bugs, ideas, or general feedback.
By agent-sh
Master workflow orchestrator with autonomous task-to-production automation, quality gates, and multi-agent review
npx claudepluginhub agent-sh/next-task --plugin next-taskValidate task completion and approve for shipping. Can be used standalone or called by the workflow. Runs autonomous validation checks.
Master workflow orchestrator with autonomous task-to-production automation
Fix CI failures and PR comments. Use this agent when ci-monitor detects issues that need code changes.
Monitor CI status and PR comments with sleep/check loops. Use this agent after PR creation to watch for issues and delegate fixes to ci-fixer.
Deep codebase analysis for understanding task context. Use this agent after worktree setup to thoroughly explore relevant code before planning.
Execute approved implementation plans with high-quality code. Use this agent after plan approval to write production-ready code following the approved plan.
Design detailed implementation plans for tasks. Create comprehensive step-by-step plans and output as structured JSON. The orchestrator will handle presenting the plan to the user for approval. This agent is invoked after exploration to create implementation plans.
Uses power tools
Uses Bash, Write, or Edit tools
Share bugs, ideas, or general feedback.
Own this plugin?
Verify ownership to unlock analytics, metadata editing, and a verified badge.
Sign in to claimOwn this plugin?
Verify ownership to unlock analytics, metadata editing, and a verified badge.
Sign in to claimBased on adoption, maintenance, documentation, and repository signals. Not a security audit or endorsement.
AI-Driven Engineering workflow commands for managing issues, tasks, implementation, and PRs.
Zero-dependency planning + execution with .flow/ task tracking and Ralph autonomous mode (multi-model review gates). Worker subagent per task for context isolation. Prime assesses 8 pillars (48 criteria) with GitHub API integration. Includes 20 subagents, 11 commands, 16 skills.
Autonomous GitHub Project Management workflow with orchestrator-agent coordination for parallel task execution via git worktrees
Smart task routing: classifies specs and routes to feature-dev or subagent-driven-dev
Install and manage TaskYou — a personal task management system with Kanban board, background AI execution, and git worktree isolation.
Harness-native ECC plugin for engineering teams - 64 agents, 262 skills, 84 legacy command shims, reusable hooks, rules, MCP conventions, and operator workflows for Claude Code plus adjacent agent harnesses
Deep repository analysis to realign project plans with actual code reality - discovers drift, gaps, and produces prioritized reconstruction plans
Plugin structure and tool use analyzer - validates plugin.json, MCP tools, and security patterns
AI slop cleanup with minimal diffs and behavior preservation
Multi-agent iterative code review until zero issues remain
Unified static analysis via agent-analyzer - git history, AST symbols, project metadata, and doc-code sync
Autonomous workflow orchestrator that takes a task from discovery to production through 12 coordinated phases - exploration, planning, implementation, multi-agent review, and shipping.
Manual development workflows leak time at every handoff: finding the right task, setting up isolated branches, remembering to run tests, reviewing your own code, syncing docs, creating PRs. next-task removes all of that. You select a task and approve a plan; everything after that runs autonomously.
Use cases:
agentsys install next-task
Requires agentsys runtime.
/next-task # Start fresh - select source, pick task, approve plan, ship
/next-task --resume # Resume active workflow
/next-task --resume 123 # Resume by task ID
/next-task --resume feature/my-task-123 # Resume by branch name
/next-task --status # Show current workflow state
/next-task --abort # Cancel and clean up
/next-task bug # Filter to bug-labeled issues only
/next-task --base develop # Target a non-default branch
The workflow has 12 phases. Phases 1-6 involve the user; phases 7-12 run autonomously.
| Phase | Name | Agent | Model |
|---|---|---|---|
| 1 | Policy Selection | - | - |
| 2 | Task Discovery | task-discoverer | Sonnet |
| 3 | Worktree Setup | worktree-manager | Haiku |
| 4 | Exploration | exploration-agent | Sonnet |
| 5 | Planning | planning-agent | Opus |
| 6 | User Approval | - | - |
| 7 | Implementation | implementation-agent | Opus |
| 8 | Pre-Review Gates | deslop:deslop-agent + prepare-delivery:test-coverage-checker | Sonnet |
| 9 | Review Loop | 4+ parallel reviewers | Sonnet |
| 10 | Delivery Validation | prepare-delivery:delivery-validator | Sonnet |
| 11 | Docs Update | sync-docs:sync-docs-agent | Sonnet |
| 12 | Ship | ship:ship | - |
Human interaction happens exactly three times: source/priority selection (Phase 1), task selection (Phase 2), and plan approval (Phase 6). Everything after Phase 6 is autonomous.
Task discovery (Phase 2) fetches from GitHub Issues, GitHub Projects, GitLab, or local markdown files. Issues with open PRs are automatically excluded. Tasks are scored by priority labels, severity, and age.
Worktree isolation (Phase 3) creates a git worktree per task so multiple workflows can run in parallel without conflicts.
Multi-agent review (Phase 9) spawns 4 core reviewers in parallel - code quality, security, performance, and test coverage - plus conditional specialists (database, API, frontend, backend, devops, architecture) based on which files changed. Runs up to 5 iterations with stall detection.
Pre-review gates (Phase 8) run deslop (AI slop cleanup) and test coverage checks in parallel before the review loop starts.
| Source | How it works |
|---|---|
| GitHub Issues | gh issue list with label-based scoring |
| GitHub Projects | gh project item-list (v2 boards) |
| GitLab Issues | glab issue list |
| Local file | Scans PLAN.md, tasks.md, TODO.md for unchecked items |
| Custom | CLI, MCP, or Skill tool |
Workflow state survives session restarts via two files:
{stateDir}/tasks.json - active task registry (in main repo){stateDir}/flow.json - workflow progress (in worktree)The state directory is platform-aware: .claude/, .opencode/, or .codex/.
| Skill | Purpose |
|---|---|
| discover-tasks | Fetch, filter, score, and present tasks for selection |
Phases 8-10 use agents from the prepare-delivery plugin:
prepare-delivery:test-coverage-checker (Phase 8), prepare-delivery:delivery-validator (Phase 10).
gh) for GitHub sources, or GitLab CLI (glab) for GitLab| Plugin | Used in |
|---|---|
| deslop | Phase 8 - AI slop cleanup |
| prepare-delivery | Phases 8-10 - test coverage, review orchestration, delivery validation |
| sync-docs | Phase 11 - documentation sync |
| ship | Phase 12 - PR creation, CI, merge |