Tasks

Skill pooyagolchian/specforge/plugins/specforge/skills/tasks

Spec-driven AI development lifecycle for Claude Code — spec-kit + AI-DLC fused, with hook-enforced human-approval gates. Installable plugin + marketplace.

Install

npx -y skills add pooyagolchian/specforge --skill tasks

Assembled from the repository path, not quoted from the project. Check it against their README if it does not work.

2 things to look at

13 days oldThe repository was created 13 days ago. New is not bad, but a brand new repository carrying a familiar-sounding name is the shape a typosquat arrives in, and there has been no time for anyone else to find a problem with it.
0 stars0 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.

What its author says it does

Copied from the file, not written here

Break an approved plan into a concrete, ordered, verifiable task list. Use when the user wants to generate tasks, create a task breakdown, turn the plan into a to-do list, or prepare for implementation. Writes tasks.md with parallel-safe [P] tags and requirement traceability.

SKILL.md

2.2 KB, as published. Nobody here has run it

/specforge:tasks — turn the plan into an executable checklist

Derive small, ordered, individually verifiable tasks from plan.md and spec.md. Each task traces to the requirement(s) it satisfies so nothing is missed.

When to use

A plan exists and the user is ready to enumerate the work.
Trigger words: "tasks", "break down", "task list", "to-dos", "prepare implementation".

Steps

Read the active feature's plan.md and spec.md.
Write tasks.md from ${CLAUDE_SKILL_DIR}/../_shared/tasks-template.md (or the skeleton below):
- Group into phases: Setup → Tests first → Implementation → Integration & polish.
- Give each task a stable ID (T001…), keep it small and verifiable, and append → FR-xxx.
- Tag tasks with no dependency on other unchecked tasks [P] (parallel-safe).
Add a Traceability check table mapping every FR/NFR to at least one task. Any unmapped requirement is a gap — fix the tasks or flag it.
Update specforge/state.md: add/ensure | NNN-slug | tasks | done | <date> |.
Audit: | <today ISO-8601> | NNN-slug | tasks | <n> tasks generated |.
Suggest /specforge:analyze for a consistency pass, then /specforge:implement (which is gated on spec + plan approval).

Tasks skeleton (if the shared template is unavailable)

# Task List: <name>
- Feature ID: NNN · Plan: ./plan.md
## Phase A — Setup
- [ ] T001 … → (structure)
## Phase B — Tests first
- [ ] T010 [P] test for US1 → FR-001
## Phase C — Implementation
- [ ] T020 implement … → FR-001
## Phase D — Integration & polish
- [ ] T030 wire end-to-end → FR-001, FR-002
## Traceability check  (every FR/NFR → task)

Guardrails

Tasks come only from the plan and spec — do not invent scope. If the plan is missing detail, return to /specforge:plan.
Prefer a tests-first ordering so acceptance criteria are encoded before implementation.