By iheanyi
Manage project tasks via CLI: add prioritized tasks with blockers/subtasks/tags, time-track with start/stop, list ready ones sorted by priority, complete with guided learnings capture. Record decisions/insights scoped to files, reflect on git changes to extract patterns, promote to docs like CLAUDE.md.
Create a new task
Mark a task as blocked by other tasks
Complete a task with guided learning capture. Use after finishing work to mark done, document learnings, and see what's unblocked.
Get full project context including tasks, learnings, and decisions
Record an architectural decision with reasoning
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.
Agent-first task management for codebases. Designed for AI agents working alongside humans.

go install github.com/iheanyi/tasuku/cmd/tk@latest
Or build from source:
git clone https://github.com/iheanyi/tasuku.git
cd tasuku && go build -o tk ./cmd/tk
sudo mv tk /usr/local/bin/
cd your-project
tk init # Create .tasuku/ directory
tk mcp install # Enable tk_* tools in your AI editor
tk hooks install # Context at session start, reminders at end
git add .tasuku/ # Tasks travel with your code
Works with Claude Code, Cursor, Copilot CLI, Codex, and OpenCode. Restart your editor after installing.
MCP Tools (40+ tools, see docs/mcp.md):
tk_add / tk_list / tk_done - Task managementtk_learn / tk_decide - Knowledge capturetk_context - Load full project stateSlash Commands (Claude Code with plugin):
/tasuku:pickup - Select and start a task/tasuku:complete - Mark done, capture learnings/tasuku:reflect - Extract learnings from recent worktk task list # See all tasks
tk task add "New feature" # Add a task
tk task start <id> # Start working
tk task done <id> # Mark complete
tk learn "insight" # Record a learning
tk ui # Terminal UI (j/k to navigate, ? for help)
| Command | Description |
|---|---|
tk task list | List tasks (--status, --tag, --tree) |
tk task add "desc" | Add task (--id, --parent, --priority) |
tk task start/done/block <id> | Lifecycle management |
tk task archive <id> | Archive completed task |
tk task restore <id> | Restore from archive |
tk learn "insight" | Record a learning |
tk decide --id X --chose Y --over Z --because "..." | Record decision |
tk task claim/release <id> | Multi-agent coordination |
tk task timer start/stop <id> | Time tracking |
tk plugin install | Install slash commands to AI tools |
tk rules sync | Sync learnings to editor rules |
Full reference: docs/cli.md
Tasks are Markdown files with YAML frontmatter:
.tasuku/
├── tasks/task-id.md # One file per task
├── archive/ # Archived tasks
├── context/
│ ├── learnings.md # Recorded insights
│ └── decisions.md # Architectural decisions
└── index.json # Auto-generated for fast queries
tk migrate beads # Convert .beads/ to .tasuku/
tk migrate beads --dry-run # Preview first
Detailed docs in docs/:
cli.md - Complete CLI commandsmcp.md - MCP tools and TUI keybindingshooks.md - Hook configurationplugins.md - Slash commands for AI toolsstorage.md - Storage formatgo test ./... # Run tests
go test -race ./... # With race detector
go build -o tk ./cmd/tk # Build
MIT
npx claudepluginhub iheanyi/tasuku --plugin tasukuAI-supervised issue tracker for coding workflows. Manage tasks, discover work, and maintain context with simple CLI commands.
Cotask — task management with TASKS.md kanban dashboard
Task management skill for coordinating complex agent work
A task management system for ambitious AI-driven development that doesn't overwhelm and confuse Cursor.
Configure Claude Code to track work using Beans.
AI-native task management system. Track progress, coordinate multi-agent work, manage hierarchical tasks with dependencies, sprints, acceptance criteria, and a web dashboard.