Skip to content
Back to skills

Task Workflow

ASecurity

How to use builtin task tools for private lightweight work tracking

  • 21 stars
  • 0 votes
  • 0 copies
  • 2 views
  • Added September 2, 2026
ai-agents

Works with

  • terminal

Security analysis

A100/100

Scanned September 2, 2026

npx -y skills add lukacf/meerkat --skill task-workflow --agent claude-code

Installs into .claude/skills of the current project.

Are you the author of Task Workflow?

Add the live security badge to your README. It updates with every re-scan.

Security grade badge for Task Workflow
[![Security: A — Skills Directory](https://www.skillsdirectory.com/api/skills/lukacf-task-workflow/badge)](https://www.skillsdirectory.com/skills/lukacf-task-workflow)

More formats (shields.io, HTML) on the badges page. Keep it an A: scan every change in CI with Pro.

Download with Pro
SKILL.md
---
name: Task Workflow
description: How to use builtin task tools for private lightweight work tracking
requires_capabilities: [builtins]
---

# Task Workflow

Use builtin task tools for lightweight project work tracking. A configured
store may persist them, but they remain scratch structure rather than the
claim-safe shared WorkGraph authority.

## Operating Rules

- Use `task_create` for planning, checklists, or local project tasks that do
  not need realm-wide claims, evidence, or cross-agent readiness.
- Use clear imperative subjects and descriptions with acceptance criteria.
- Read an exact task with `task_get`, and use `task_update` to move it to
  `in_progress` when you start and `completed` only when it is actually done.
- Use `task_list` after completing a task to find the next local item.
- Keep dependencies simple. If blocking relationships must coordinate multiple
  agents or survive compaction/restarts as shared truth, use WorkGraph instead.

## Boundary With WorkGraph

- Builtin tasks: lightweight project task lists and simple progress tracking;
  no claims, leases, revisions, or evidence authority.
- WorkGraph: realm-scoped durable commitments, readiness, dependency topology,
  authorized claims, leases, evidence, and terminal truth. Its ledger is the
  shared authority, not a peer message, prompt, actor, or task-list projection.
- Schedule: time-based wakeups, recurrence, occurrence claims, and delivery
  outcomes. It does not decide WorkGraph readiness or completion.
- Memory: recalled knowledge, not live task state.

Attribution

Is this your skill, or is something wrong with this listing? Request removal or report an issue. Author removals are honored within 72 hours.

Comments

Loading comments…

Newsletter

Get the weekly list of safe new skills, and the ones we pulled.