Standardmäßig ist der Prompt ausgewählt, der zuerst die Quelle prüft. Sie können zu einem direkten Befehl wechseln oder eine lokale Kopie herunterladen.
Quelldateien prüfen
Lesen Sie SKILL.md und alle von SkillsMP angezeigten Begleitdateien, bevor Sie sich für eine Installation entscheiden.
Mit Codex oder Claude installieren Kopieren Sie diesen Prompt, fügen Sie ihn in Codex, Claude oder einen anderen Assistant ein und lassen Sie die Skill-Seite prüfen und installieren.
Ein direkter Befehl überspringt den Prüf-Prompt. Prüfen Sie die Quelle, bevor Sie ihn ausführen.
AI-powered task management CLI for project initialization, PRD parsing, task breakdown, and execution with multi-provider AI support
Task-O-Matic
AI-powered task management CLI for single projects. Helps initialize projects, parse PRDs into tasks, break down work with AI, and execute tasks with automatic retry and verification.
When to Use
Use when the user wants to:
Initialize a new project with task-o-matic
Attach task-o-matic to an existing project (detect stack automatically)
Parse a PRD (Product Requirements Document) into structured tasks
Break down tasks into smaller subtasks using AI
Enhance tasks with AI-powered documentation research
Execute tasks with AI coding assistants
Configure AI providers and models
CLI Location
The CLI binary is at: packages/cli/dist/cli/bin.js (after building)
Run with: node packages/cli/dist/cli/bin.js <command> [options]
Or if installed globally: task-o-matic <command> [options]
# Attach to existing project (auto-detect stack)
task-o-matic init attach
# With full project analysis
task-o-matic init attach --analyze
# Dry run (just show detection, don't create files)
task-o-matic init attach --dry-run
# Force re-detection (update cached stack.json)
task-o-matic init attach --redetect
Options:
--analyze: Run full project analysis (TODOs, features, structure)
--dry-run: Just detect stack, don't create files
--redetect: Force re-detection (overwrites cached stack.json)
--ai-provider <provider>: AI provider (openrouter/anthropic/openai/custom)
--ai-model <model>: AI model
--context7-api-key <key>: Context7 API key
What it detects:
Language (TypeScript/JavaScript)
Framework(s) (Next.js, Express, Hono, etc.)
Database (Postgres, MongoDB, SQLite, etc.)
ORM (Prisma, Drizzle, etc.)
Auth (Better-Auth, Clerk, NextAuth, etc.)
Package Manager (npm, pnpm, bun, yarn)
Runtime (Node, Bun)
API style (tRPC, GraphQL, REST)
Testing frameworks
Build tools
What it creates:
.task-o-matic/config.json - AI settings
.task-o-matic/stack.json - Cached stack detection (used by AI context)
.task-o-matic/mcp.json - Context7 config
.task-o-matic/tasks/ - Task storage
.task-o-matic/prd/ - PRD storage
Updates .gitignore if git exists
Configure AI Provider
# Set provider and model
task-o-matic config set-ai-provider openrouter xiaomi/mimo-v2-flash:free
# Set with custom URL
task-o-matic config set-ai-provider custom https://api.example.com
# Check current config
task-o-matic config get-ai-config
Generate PRD from Codebase
# Analyze current project and generate PRD
task-o-matic prd generate
# With streaming output
task-o-matic prd generate --stream
# With custom output filename
task-o-matic prd generate --output my-project-prd.md
Parse PRD
# Basic parse
task-o-matic prd parse --file prd.md
# With streaming output
task-o-matic prd parse --file prd.md --stream
# With reasoning tokens (for models that support it)
task-o-matic prd parse --file prd.md --ai-reasoning 4000 --stream
# Multiple models with combination
task-o-matic prd parse --file prd.md --ai openrouter:model1 openrouter:model2 --combine-ai openrouter:combine-model
Options:
--file <path>: Path to PRD file (required)
--stream: Show streaming AI output
--ai-reasoning <tokens>: Max reasoning tokens for supported models
--ai <models...>: Multiple AI models
--combine-ai <provider:model>: Model to combine results
--tools: Enable filesystem tools for project analysis
Work with Tasks
List Tasks
task-o-matic tasks list
# With filters
task-o-matic tasks list --status todo
task-o-matic tasks list --tag frontend
# Split specific task
task-o-matic tasks split --task-id 1 --stream
# Split all tasks
task-o-matic tasks split --all --stream
# With filters
task-o-matic tasks split --all --status todo --stream
Options:
--task-id <id>: Specific task to split
--all: Split all tasks without subtasks
--status <status>: Filter by status (todo/in-progress/completed)
--stream: Show streaming output
--ai-reasoning <tokens>: Enable reasoning
--dry: Preview without changes
Enhance Tasks
# Enhance specific task
task-o-matic tasks enhance --task-id 1 --stream
# Enhance all tasks
task-o-matic tasks enhance --all --stream
# With filter
task-o-matic tasks enhance --all --status todo --stream
Options:
--task-id <id>: Specific task to enhance
--all: Enhance all tasks
--status <status>: Filter by status
--tag <tag>: Filter by tag
--dry: Preview without changes
--force: Skip confirmation
Task Status
# Set task status
task-o-matic tasks status --task-id 1 --status in-progress
# Available statuses: todo, in-progress, completed
Task Tree
# Show hierarchical task tree
task-o-matic tasks tree
Get Next Task
# Get next task to work on
task-o-matic tasks get-next
Execute Tasks
# Execute all todo tasks
task-o-matic tasks execute-loop --status todo
# Execute specific tasks
task-o-matic tasks execute-loop --ids 1,2,3
# With specific tool
task-o-matic tasks execute-loop --tool claude
# With planning phase
task-o-matic tasks execute-loop --plan --review-plan
# With verification
task-o-matic tasks execute-loop --verify "bun run test" --verify "bun run type-check"# Progressive model escalation
task-o-matic tasks execute-loop --try-models "gpt-4o-mini,gpt-4o,claude:sonnet-4"
