Skip to content
Back to skills

690 Argument Skill 1ddab9e3

ASecurity

SAM-style feature initiation workflow: discovery -> codebase analysis -> architecture spec -> task decomposition -> validation -> context manifest. Use when a user asks to add a feature, plan a feature, or convert an idea into executable task files.

  • 9 stars
  • 0 votes
  • 0 copies
  • 0 views
  • Added October 11, 2026
testingpythongotesting

Works with

  • cli

Security analysis

A100/100

Scanned October 11, 2026

npx -y skills add tools-only/X-Skills --skill 690-argument-skill_1ddab9e3 --agent claude-code

Installs into .claude/skills of the current project.

Are you the author of 690 Argument Skill 1ddab9e3?

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

Security grade badge for 690 Argument Skill 1ddab9e3
[![Security: A — Skills Directory](https://www.skillsdirectory.com/api/skills/tools-only-690-argument-skill-1ddab9e3/badge)](https://www.skillsdirectory.com/skills/tools-only-690-argument-skill-1ddab9e3)

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

SKILL.md
---
argument-hint: <feature description or existing doc path>
user-invocable: true
description:"SAM-style feature initiation workflow: discovery -> codebase analysis -> architecture spec -> task decomposition -> validation -> context manifest. Use when a user asks to add a feature, plan a feature, or convert an idea into executable task files."
version: 1.0.0
last_updated: '2026-01-27'
python_compatibility: 3.11+
---

# Add New Feature (SAM Workflow)

You MUST convert the user's request into **durable SAM artifacts** under the repo:

- `plan/feature-context-{slug}.md` (discovery)
- `plan/codebase/{FOCUS}.md` (optional, analysis)
- `plan/architect-{slug}.md` (architecture/design spec)
- `plan/tasks-{N}-{slug}.md` (executable task plan with Agents, deps, and verification)

<feature_request>
$ARGUMENTS
</feature_request>

---

## Orchestrator Discipline

You are an orchestrator. You coordinate work across specialized agents. Prefer delegating discovery and analysis.

---

## Phase 1: Discovery (feature-researcher)

Delegate to `feature-researcher` to produce `plan/feature-context-{slug}.md` and questions for resolution.

---

## Phase 2: Codebase Analysis (codebase-analyzer)

If helpful, delegate to `codebase-analyzer` for one or more focus areas:

- patterns
- architecture
- testing
- conventions

Outputs go to `plan/codebase/`.

---

## Phase 3: Architecture Spec (python-cli-design-spec)

Delegate to `python-cli-design-spec` to write `plan/architect-{slug}.md` based on:

- the feature context doc
- codebase analysis docs (if created)
- existing repo constraints (`CLAUDE.md`, `pyproject.toml`, etc.)

---

## Phase 4: Task Decomposition (swarm-task-planner)

Delegate to `swarm-task-planner` to:

- create `plan/tasks-{N}-{slug}.md`
- ensure every task has:
  - **Status**, **Dependencies**, **Priority**, **Complexity**, **Agent**
  - Acceptance Criteria (3+)
  - Verification Steps (3+)

---

## Phase 5: Plan Validation Gate (plan-validator)

Delegate to `plan-validator`. If it returns `BLOCKED`, do not proceed.

---

## Phase 6: Context Manifest (context-gathering)

Delegate to `context-gathering` with the task file path. It must insert a `## Context Manifest` into the task file.

---

## Success Outcome

When all phases complete, provide the user:

- the feature slug
- the task file path
- next step: run the `implement-feature` skill with the slug or task file path

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.