agentskills.codes
TA

Generate concise but complete instructions that teach any model how to use every Taskery feature across CLI, API, and web board workflows.

Install

mkdir -p .claude/skills/taskery && curl -L -o skill.zip "https://agentskills.codes/api/skills/download/16518" && unzip -o skill.zip -d .claude/skills/taskery && rm skill.zip

Installs to .claude/skills/taskery

Activation

This is the description your AI agent reads to decide when to run this skill — the better it matches your request, the more reliably it fires.

Generate concise but complete instructions that teach any model how to use every Taskery feature across CLI, API, and web board workflows.
138 charsno explicit “when” trigger

About this skill

Taskery Full Operator

Use this skill when the user wants a reusable instruction set/prompt that teaches another model how to operate Taskery end-to-end.

Defaults

Unless the user specifies otherwise, assume:

  • API base URL: http://127.0.0.1:4010
  • Web board URL: http://127.0.0.1:3010
  • Preferred CLI: taskery <command>
  • Source-repo CLI fallback: pnpm --filter taskery exec node --import tsx src/bin/taskboard.ts <command>

Output Rules

Return one concise Markdown guide that is still complete. Keep it practical and command-first.

You must cover all of the following facets:

  • Core lifecycle: create, inspect, update, move across statuses, review, complete, delete
  • CLI commands: create, list, show, update, move, delete, settings, up
  • CLI filters/flags, including --status, --priority, --assignee, --titleContains, and notification settings flags
  • Optimistic concurrency with expectedVersion (required for move and delete, recommended for update)
  • Conflict handling (VERSION_CONFLICT => re-fetch and retry once)
  • Status and priority enums
  • JSON-first behavior and exit codes (0..4)
  • API route surface and payload shapes
  • Web board features (drag/drop, edit modal, due dates, sync indicator, notification settings)
  • Environment variables and run-mode differences (taskery up vs API/web dev split)

Required Structure In Generated Guide

Use these section headings in order:

  1. Taskery At A Glance
  2. Operating Rules For Models
  3. CLI Commands
  4. Task Lifecycle Playbooks
  5. Notification Settings
  6. API Surface
  7. Web Board Features
  8. Run Modes And Env
  9. Failure Handling

Canonical Facts To Include

  • Statuses: PENDING | STARTED | BLOCKED | REVIEW | COMPLETE
  • Priorities: LOW | MEDIUM | HIGH | URGENT
  • CLI defaults to JSON output; --text is optional for humans
  • CLI exit codes:
    • 0: success
    • 1: internal/runtime error
    • 2: validation error
    • 3: task not found
    • 4: version conflict
  • API endpoints:
    • GET /api/health
    • GET /api/tasks
    • POST /api/tasks
    • PATCH /api/tasks/:id
    • POST /api/tasks/:id/move
    • DELETE /api/tasks/:id
    • GET /api/settings/notifications
    • PATCH /api/settings/notifications
  • API response shapes:
    • GET /api/tasks => { tasks, notificationSettings }
    • task mutations => { task }
    • settings routes => { settings }
    • errors => { code, message, details? }

Instruction Quality Bar

  • Keep it concise, but do not omit any feature category above.
  • Prefer executable examples over prose.
  • Do not invent unsupported endpoints or commands.
  • For any write operation, instruct models to verify state afterward with list or show.
  • State explicitly that show obtains the current version for safe writes.

Search skills

Search the agent skills registry