By mnthe
Native teammate-based collaboration plugin with role-based workers, event-driven coordination, and Claude Code agent teams.
Clean up teamwork project
Show teamwork project status dashboard
Manually trigger project verification
Join a teamwork project as a worker teammate
Start a teamwork project with native agent team collaboration
Backend specialist worker for teamwork. API, services, database, business logic. <example> Context: Orchestrator spawns a backend worker for API and database tasks user: (spawned by orchestrator via Task()) assistant: Checks TaskList for backend tasks, claims API endpoint task, implements endpoint with database integration and validation, runs tests, collects evidence (curl output, test results with exit codes), marks task completed, reports to orchestrator via SendMessage </example>
DevOps specialist worker for teamwork. CI/CD, deployment, infrastructure, containerization. <example> Context: Orchestrator spawns a devops worker for infrastructure tasks user: (spawned by orchestrator via Task()) assistant: Checks TaskList for devops tasks, claims Docker configuration task, creates Dockerfile with multi-stage build, adds docker-compose.yml, tests build succeeds, collects evidence (build logs, container runs with exit codes), marks completed, reports to orchestrator via SendMessage </example>
Documentation specialist worker for teamwork. README, API docs, examples, guides. <example> Context: Orchestrator spawns a docs worker for documentation tasks user: (spawned by orchestrator via Task()) assistant: Checks TaskList for docs tasks, claims API documentation task, reads implemented API endpoints, creates comprehensive API.md with endpoint documentation, verifies examples match implementation, collects evidence (file created, examples tested with exit codes), marks completed, reports to orchestrator via SendMessage </example>
Final project verifier for teamwork. Reads all tasks, verifies evidence, runs full build and tests. <example> Context: Orchestrator spawns final-verifier after all tasks are completed user: (spawned by orchestrator via Task()) assistant: Reads all tasks via TaskList, verifies each task has status completed with concrete evidence, scans changed files for blocked patterns, runs full project build and test suite, sends PASS/FAIL verdict to orchestrator via SendMessage with detailed findings </example>
Frontend specialist worker for teamwork. UI, components, styling, user interactions. <example> Context: Orchestrator spawns a frontend worker for UI component tasks user: (spawned by orchestrator via Task()) assistant: Checks TaskList for frontend tasks, claims component creation task, implements React component with TypeScript and styling, verifies component renders without errors, collects evidence (file created, build passes, tests pass with exit codes), marks completed, reports to orchestrator via SendMessage </example>
Event-driven coordination for teamwork orchestrators. Replaces polling-based monitoring loop with TeammateIdle and TaskCompleted hooks.
Hybrid task decomposition strategy for teamwork projects. Covers plan-based and semantic decomposition, granularity rules, role assignment, dependency management, and complexity assessment.
Reset and clean up teamwork project state. Use when project is stuck, failed, or needs a fresh start. Uses native TeamDelete API for cleanup.
8-phase task execution lifecycle for teamwork workers using native Claude Code APIs. Covers task discovery, claiming, structured description parsing, TDD workflow, implementation, verification, selective commit, and completion reporting.
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.
Claude Code plugin marketplace focused on "hardworker" productivity patterns: verification-first development, multi-session collaboration, and evidence-based completion.
Quick Install:
claude plugin marketplace add mnthe/hardworker-marketplace
claude plugin install ultrawork@hardworker-marketplace
Detailed Comparison → | Architecture →
# Add marketplace
claude plugin marketplace add mnthe/hardworker-marketplace
# Install verification-first development
claude plugin install ultrawork@hardworker-marketplace
# Install multi-session collaboration
claude plugin install teamwork@hardworker-marketplace
# Install knowledge extraction
claude plugin install knowledge-extraction@hardworker-marketplace
# Start verification-first development session
/ultrawork "implement user authentication with JWT"
# System creates plan with tasks and success criteria
# Worker agents execute tasks in parallel
# Verifier checks all criteria with concrete evidence
# Check status
/ultrawork-status
# Start coordination (orchestrator spawns native teammates automatically)
/teamwork "build REST API with tests and docs"
# Orchestrator creates tasks, spawns role-based workers as native teammates
# Workers execute tasks, coordinate via SendMessage
# Event-driven: TaskCompleted and TeammateIdle hooks drive progress
# Check status
/teamwork-status
| Scenario | Recommendation |
|---|---|
| Solo developer, strict verification | ultrawork |
| Team collaboration, parallel work | teamwork |
| Capturing code patterns and insights | knowledge-extraction |
| CI/CD and automation | ultrawork --auto |
| Parallel agent collaboration | teamwork |
| TDD workflow enforcement | ultrawork |
See Detailed Scenarios below for more guidance.
Strict verification-first development mode with session isolation and mandatory evidence collection.
Key Features: Session isolation, mandatory planning with success criteria, evidence-based completion, Execute→Verify loop with auto-retry, parallel task execution, cross-platform support.
Role-based worker agents for parallel development using Claude Code's native teammate API.
Key Features: Role-based worker agents (frontend, backend, devops, test, docs, security, review), native teammate API, event-driven coordination (TaskCompleted, TeammateIdle hooks), dashboard status, project and team isolation.
Extract and manage knowledge from codebases for AI agent context.
Key Features: Pattern-based knowledge extraction, markdown documentation generation, context management for AI agents, integration with ultrawork and teamwork.
| Feature | ultrawork | teamwork | knowledge-extraction |
|---|---|---|---|
| Primary Use Case | Solo developer, strict verification | Agent team collaboration, parallel work | Pattern capture, documentation |
| Session Model | Single session, isolated worktree | Single session, native teammate API | Session-based insights |
| Verification | Mandatory, multi-tier (task + final) | Final verifier agent | Not applicable |
| Evidence Collection | Automatic via hooks | Structured via workers | Insight extraction |
| Concurrency | Parallel workers, single session | Native teammates, event-driven coordination | Single session |
| Role Specialization | No | Yes (8 worker roles) | No |
| Auto Mode | Yes (--auto flag) | No (interactive only) | No (hook-based) |
| Best For | Feature implementation, bug fixes | Large features, parallel agent work | Learning, documentation |
flowchart TD
CLI[Claude Code CLI] --> Commands
CLI --> Hooks[Lifecycle Hooks]
npx claudepluginhub mnthe/hardworker-marketplace --plugin teamworkCollect insights from Claude sessions and extract them into reusable components (Skills, Commands, Agents, CLAUDE.md, Rules Files).
Strict verification-first development mode with Session ID-based isolation (Bun cross-platform version).
Dynamically assemble expert agent teams for complex tasks using Claude Code's agent teams feature
AI team role and worker manager for multi-agent development workflows.
Multi-agent orchestration for Claude Code. 12 specialized agents working in parallel — planning, building, reviewing, debugging. Plus a Hub for always-alive multi-project sessions controllable from Telegram or Slack.
Multi-agent team orchestration for Claude Code. Set up parallel AI agent teams with file-based planning, progress tracking, and role-based collaboration.
General-purpose agent superteam for Claude code - contract-gated verification loops with 7 outer-loop roles, 5-phase orchestration, task form-driven inner-loop cooperation, and structural prevention of premature exits.
Orchestrate multi-agent teams for parallel code review, hypothesis-driven debugging, and coordinated feature development using Claude Code's Agent Teams