speckit-companion-tasks

Generates dependency-ordered task lists organized by user story from spec-kit plans.

Updated May 28, 2026
One-click install
npx skills add https://github.com/whw23/PingoGate --skill speckit-companion-tasks-whw23
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: speckit-companion-tasks
Source: https://github.com/whw23/PingoGate/tree/main/.claude/skills/speckit-companion-tasks
Command: npx skills add https://github.com/whw23/PingoGate --skill speckit-companion-tasks-whw23

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? Turning a feature specification and implementation plan into an actionable, dependency-ordered task list is tedious and error-prone. This Skill automates that step inside a spec-kit project, producing a tasks.md file where every task maps to a user story and an exact file path. ## Core Features & Use Cases - User-story phased task generation: Reads plan.md, spec.md, and optional design artifacts to produce tasks.md grouped into Setup, Foundational, per-story, and Polish phases. - Explicit dependency waves: Groups independent tasks into parallelizable waves marked with [P], with join lines showing where work must wait. - Size-aware output: Reads the recorded change size from .spec-context.json and trims ceremony for simple changes while keeping full structure for normal or oversized ones. - Use Case: After running the plan step on a new feature, invoke this Skill to generate a tasks.md where each story is an independently testable increment ready for the implement step. ## Quick Start Generate the phased task list for the current spec-kit feature from its plan and spec.

Frequently Asked Questions about speckit-companion-tasks

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I generate a task list from a spec-kit plan?▼

Run this command inside a spec-kit project after the plan step. It reads plan.md and spec.md from the feature directory recorded in .specify/feature.json and writes a tasks.md organized by user story into dependency-ordered phases.

How are tasks ordered and parallelized in the generated task list?▼

Tasks are grouped into waves of independent work touching different files, each tagged [P], with explicit join lines showing where later waves must wait. Phases run Setup, Foundational, one phase per user story, then Polish.

Does this work without a spec-kit project structure?▼

No. It requires a spec-kit project with a .specify directory containing feature.json, plus plan.md and spec.md in the feature directory. Optional inputs like data-model.md, contracts, and research.md are used when present.

What changes for small or simple features?▼

When .spec-context.json records the size as simple, the Skill produces a lean list: no baseline setup task, no per-story goal or checkpoint blocks, and a one-line dependency note instead of the full execution-order prose.

Why should I not edit .spec-context.json or tasks.md checkboxes by hand?▼

The companion write-context.py script owns timing records and checkbox flips. Hand edits can corrupt the JSON with duplicated keys, while the script writes atomically, stamps real clock times, and stays idempotent.