{
  "markdown": "# next-task\n\nAutonomous workflow orchestrator that takes a task from discovery to production through 12 coordinated phases - exploration, planning, implementation, multi-agent review, and shipping.\n\n## Why\n\nManual 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.\n\n**Use cases:**\n- Pick the highest-priority issue from GitHub/GitLab and ship it end-to-end\n- Resume an interrupted workflow from exactly where it stopped\n- Run parallel workflows in isolated git worktrees\n\n## Installation\n\n```bash\nagentsys install next-task\n```\n\nRequires [agentsys](https://github.com/agent-sh/agentsys) runtime.\n\n## Quick Start\n\n```\n/next-task                              # Start fresh - select source, pick task, approve plan, ship\n/next-task --resume                     # Resume active workflow\n/next-task --resume 123                 # Resume by task ID\n/next-task --resume feature/my-task-123 # Resume by branch name\n/next-task --status                     # Show current workflow state\n/next-task --abort                      # Cancel and clean up\n/next-task bug                          # Filter to bug-labeled issues only\n/next-task --base develop               # Target a non-default branch\n```\n\n## How It Works\n\nThe workflow has 12 phases. Phases 1-6 involve the user; phases 7-12 run autonomously.\n\n| Phase | Name | Agent | Model |\n|-------|------|-------|-------|\n| 1 | Policy Selection | - | - |\n| 2 | Task Discovery | task-discoverer | Sonnet |\n| 3 | Worktree Setup | worktree-manager | Haiku |\n| 4 | Exploration | exploration-agent | Sonnet |\n| 5 | Planning | planning-agent | Opus |\n| 6 | User Approval | - | - |\n| 7 | Implementation | implementation-agent | Opus |\n| 8 | Pre-Review Gates | deslop:deslop-agent + prepare-delivery:test-coverage-checker | Sonnet |\n| 9 | Review Loop | 4+ parallel reviewers | Sonnet |\n| 10 | Delivery Validation | prepare-delivery:delivery-validator | Sonnet |\n| 11 | Docs Update | sync-docs:sync-docs-agent | Sonnet |\n| 12 | Ship | ship:ship | - |\n\n**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.\n\n### Phase highlights\n\n**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.\n\n**Worktree isolation** (Phase 3) creates a git worktree per task so multiple workflows can run in parallel without conflicts.\n\n**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.\n\n**Pre-review gates** (Phase 8) run deslop (AI slop cleanup) and test coverage checks in parallel before the review loop starts.\n\n## Task Sources\n\n| Source | How it works |\n|--------|-------------|\n| GitHub Issues | `gh issue list` with label-based scoring |\n| GitHub Projects | `gh project item-list` (v2 boards) |\n| GitLab Issues | `glab issue list` |\n| Local file | Scans `PLAN.md`, `tasks.md`, `TODO.md` for unchecked items |\n| Custom | CLI, MCP, or Skill tool |\n\n## State Management\n\nWorkflow state survives session restarts via two files:\n\n- `{stateDir}/tasks.json` - active task registry (in main repo)\n- `{stateDir}/flow.json` - workflow progress (in worktree)\n\nThe state directory is platform-aware: `.claude/`, `.opencode/`, or `.codex/`.\n\n## Skills\n\n| Skill | Purpose |\n|-------|---------|\n| discover-tasks | Fetch, filter, score, and present tasks for selection |\n\nPhases 8-10 use agents from the [prepare-delivery](https://github.com/agent-sh/prepare-delivery) plugin:\n`prepare-delivery:test-coverage-checker` (Phase 8), `prepare-delivery:delivery-validator` (Phase 10).\n\n## Requirements\n\n- [agentsys](https://github.com/agent-sh/agentsys) runtime\n- Git 2.20+ (worktree support)\n- GitHub CLI (`gh`) for GitHub sources, or GitLab CLI (`glab`) for GitLab\n- Node.js 18+\n\n## Cross-Plugin Dependencies\n\n| Plugin | Used in |\n|--------|---------|\n| [deslop](https://github.com/agent-sh/deslop) | Phase 8 - AI slop cleanup |\n| [prepare-delivery](https://github.com/agent-sh/prepare-delivery) | Phases 8-10 - test coverage, review orchestration, delivery validation |\n| [sync-docs](https://github.com/agent-sh/sync-docs) | Phase 11 - documentation sync |\n| [ship](https://github.com/agent-sh/ship) | Phase 12 - PR creation, CI, merge |\n\n## Related Plugins\n\n- [ship](https://github.com/agent-sh/ship) - standalone PR and release workflow\n- [enhance](https://github.com/agent-sh/enhance) - plugin structure and config analysis\n- [audit-project](https://github.com/agent-sh/audit-project) - multi-agent code review (standalone)\n\n## License\n\nMIT\n",
  "bytes": 5106,
  "sha": "e86817fc5946bbc8a50a166143923635b3577d63d9cc2887cbd2529ced3fd16c",
  "repo_slug": "agent-sh/next-task",
  "fonte": "repo",
  "truncated": false,
  "api": "https://agentalog.com/api/listings/plg_agent_sh_next_task_next_task_73b9c78b/readme"
}