Skip to content
Back to skills

Using Dev Cycle

ASecurity

Use when starting any conversation in a repo with the dev-cycle plugin - routes feature, bugfix, architecture, and planning requests to the right dev-cycle skill before any other action.

  • 2 stars
  • 0 votes
  • 0 copies
  • 1 view
  • Added October 19, 2026
developmentcode-reviewfrontend

Works with

  • cli

Security analysis

A100/100

Pro scans all 2 files and shows the line behind each finding

Scanned October 19, 2026

npx -y skills add petruseugene/full-development-cycle --skill using-dev-cycle --agent claude-code

Installs into .claude/skills of the current project.

Are you the author of Using Dev Cycle?

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

Security grade badge for Using Dev Cycle
[![Security: A — Skills Directory](https://www.skillsdirectory.com/api/skills/petruseugene-using-dev-cycle/badge)](https://www.skillsdirectory.com/skills/petruseugene-using-dev-cycle)

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: using-dev-cycle
disable-model-invocation: true
description: Use when starting any conversation in a repo with the dev-cycle plugin - routes feature, bugfix, architecture, and planning requests to the right dev-cycle skill before any other action.
---

<SUBAGENT-STOP>
If you were dispatched as a subagent to execute a specific dev-cycle task, ignore
this skill and follow your dispatch prompt.
</SUBAGENT-STOP>

# Using dev-cycle

## When to use

Read at the start of every session. It decides *whether* a request belongs to the
dev-cycle pipeline and *which* skill takes it.

## Required reading

1. Read `${CLAUDE_PLUGIN_ROOT}/skills/playbook/SKILL.md` (the core) — but only once a
   dev-cycle skill has been selected below. Do not read it for out-of-scope requests.

## The job

Route the request. If it matches a row, invoke that skill BEFORE any other action —
before exploring the codebase, before clarifying questions.

| Request looks like | Route |
|---|---|
| "build X", "add feature", "implement Y end-to-end", "ship X" | Skill `dev-cycle:orchestrator` |
| "fix bug", "X is broken", "hotfix" | Skill `dev-cycle:orchestrator` (bugfix flow) |
| "how should we structure X", "which architecture", "ADR", "C4 diagram" | Skill `dev-cycle:architector` |
| "write a spec for X" (without implementing it) | `/dev-cycle:spec-writer`, or Read `${CLAUDE_PLUGIN_ROOT}/skills/spec-writer/SKILL.md` |
| "decompose X into tasks", "write a plan" (without implementing it) | `/dev-cycle:plan-writer`, or Read `${CLAUDE_PLUGIN_ROOT}/skills/plan-writer/SKILL.md` |
| "review this diff/branch" | `/dev-cycle:code-reviewer`, or Read `${CLAUDE_PLUGIN_ROOT}/skills/code-reviewer/SKILL.md` |
| "show me the feature dashboard" | `/dev-cycle:dashboard` |

Only the orchestrator and architector are invoked through the Skill tool. Standalone
routes use a slash command or read-by-path, never the Skill tool.

The phrases above are English examples. Match the request by intent, not wording:
requests in any language (e.g. Russian "добавь фичу", "исправь баг") route the same way.

NOT dev-cycle work — handle normally, do not invoke the pipeline:
- One-line edits, renames, formatting
- "How does this work" questions, reading code, explanations
- Step-by-step pair programming where the user is driving

Rules:
- The orchestrator NEVER writes feature code — it dispatches subagents.
- Individual dev-cycle skills (`dev-cycle:frontend-dev`, `dev-cycle:auto-tester`,
  `dev-cycle:cli-fallback`, ...) are dispatched BY the orchestrator, not invoked
  directly by the user — unless the user names one explicitly.
- Announce the chosen route in one line.
- User instructions (CLAUDE.md, direct requests) override this routing.

## Output contract

This skill produces no report. It only selects the next skill to invoke.

Files in this skill

  • SKILL.md2.8 KB
  • primer.md622 B

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.