taskmaster-skill
Skill22 starsChanged 7 months agoMITView on GitHub
Use when managing complex project plans, tracking multi-phase task progress, or prioritizing development tasks
- Installs packages
In this repository17 other files it gives its agents.
- code-analysisSkill
- code-generatorSkill
- contributingSkill
- daily-logSkill
- devflowSkill
- git-workflowSkill
- hexo-blog-updateSkill
- kimicode-vision-bridgeSkill
- knowledge-base-cacheSkill
- layered-first-principles-teachingSkill
- msvc-buildSkill
- mvp-designSkill
- paper-first-principlesSkill
- requirement-clarificationSkill
- research-to-practiceSkill
- ue5-umgSkill
- unity-mcpSkill
The file
We split this skill into the rules it gives an agent
3 rules. About 0 are common, 1 is distinctive, and 2 are very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈1
Distinctive · estimate
This author's own choices. Worth a look.
2
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 166
All commands follow this pattern:
- Very specific: line 183
Check dependencies before start (run deps first)
- Very specific: line 184
Verify before archive (run verify first)
≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
TaskMaster Skill
Agent-optimized task management system for AgentFabric.
Overview
A structured task management system for organizing, tracking, and executing multi-phase development plans through agent-optimized CLI workflows.
When to Use
- Managing complex project plans with interdependent tasks
- Tracking multi-phase task progress across milestones or sprints
- Prioritizing development tasks based on dependencies and readiness
- Coordinating sequential work that requires structured start, verify, and archive workflows
Do not use when:
- Simple single-step tasks without dependencies
- Informal tracking where a basic todo list is sufficient
- Projects not requiring configured plan directories or CLI tooling
Setup
Prerequisites
- Node.js 22+
- Working directory:
docs/plans/
Installation
Option 1: Direct Node.js (Agent Recommended)
cd docs/plans/taskmaster-skill/cli
npm installThen use CLI via:
node cli/task.js <command>Option 2: Shortcut Scripts (Install Dependencies Automatically)
cd docs/plans
# Windows
taskmaster-skill\task.bat <plan> <command>
# Linux/Mac
./taskmaster-skill/task.sh <plan> <command>Quick Start
# View all plans
node cli/task.js plans
# Check progress
node cli/task.js progress mvp_v1
# List tasks
node cli/task.js list mvp_v1
# List ready-to-start tasks (dependencies completed)
node cli/task.js ready mvp_v1
node cli/task.js ready mvp_v1 --limit 20Configuration (REQUIRED)
Before using TaskMaster, you MUST configure the plan mapping:
Step 1: Copy template
cp config.json.template config.jsonStep 2: Edit config.json
{
"version": "1.0.0",
"plans": {
"mvp_v1": {
"path": "E:\\Agent\\agent-fabric\\docs\\plans\\mvp_v1",
"description": "AgentFabric MVP v1 实施计划"
}
},
"settings": {
"defaultPlan": "mvp_v1",
"autoCreatePlan": false,
"logRetentionDays": 30
}
}Replace the path with your actual plan directory (absolute path).
Step 3: Verify configuration
# List configured plans
node cli/task.js config list
# Add additional plans if needed
node cli/task.js config add <name> <path> --description "<desc>"
# Set default plan
node cli/task.js config set-default <name>Documentation Index
Agent Instruction Manuals
| Document | Purpose | When to Use |
| COMMANDS.md | Complete command specifications | Need detailed syntax for any command |
| WORKFLOWS.md | Standard operation procedures | User asks to start/complete/audit tasks |
| BEST_PRACTICES.md | Operational rules and constraints | Need to validate operation against rules |
Reference Data
| Document | Purpose | When to Use |
| DATA_FORMAT.md | JSON schema specifications | Need to parse or construct data |
| TROUBLESHOOTING.md | Error codes and resolution procedures | Command returns error |
Quick Reference
| Document | Purpose | When to Use |
| EXAMPLES.md | Scenario-based execution patterns | User intent matches example scenario |
| TEMPLATES.md | Copy-paste command templates | Need ready-to-use command |
Quick Decision Map
User Request
|
+-- "查看/查询/了解 (View/Query/Understand)..." ──► Read COMMANDS.md "Query Commands" section
+-- "有哪些任务可以开始 (What tasks can I start)" ──► Read COMMANDS.md "ready" command
+-- "开始/做...任务 (Start/Do...task)" ──► Read WORKFLOWS.md "Start Task Workflow"
+-- "完成/做完了 (Complete/Done)..." ──► Read WORKFLOWS.md "Complete Task Workflow"
+-- "添加/新建 (Add/Create)...task" ──► Read COMMANDS.md "add" + EXAMPLES.md "Adding New Task"
+-- "归档 (Archive)..." ──► Verify done + verified ──► Read COMMANDS.md "archive"
+-- "检查/核验 (Check/Verify)..." ──► Read COMMANDS.md "verify"
+-- "遇到问题/报错 (Encounter problem/error)..." ──► TROUBLESHOOTING.mdCommand Syntax
All commands follow this pattern:
node cli/task.js <command> [options]Or via shortcut scripts:
# Windows
taskmaster-skill\task.bat <plan> <command>
# Linux/Mac
./taskmaster-skill/task.sh <plan> <command>Execution Rules
ALWAYS:
- Check dependencies before
start(rundepsfirst) - Verify before
archive(runverifyfirst) - Parse command output to determine next action
- Confirm with user for
--forceoperations
NEVER:
- Skip
verifybeforearchive(unless user explicitly confirms--force) - Assume task ID without confirming
- Ignore error messages in command output
File Locations
docs/plans/
├── taskmaster-skill/
│ ├── cli/task.js # CLI entry point
│ ├── config.json # Plan configurations
│ ├── logs/ # Operation logs
│ │ ├── operations.log
│ │ └── operations.jsonl
│ └── *.md # This documentation
│
└── <plan-name>/
├── data.json # Task data (JSON format)
└── README.md # Plan descriptionVersion
- CLI Version: 1.1.0
- Data Format: 1.0.0
- Last Updated: 2026-03-04
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

