What is the Spectre workflow?

Spectre stands for Scope, Plan, Execute, Clean, Test, Rebase, and Evaluate. It is a structured sequence designed to provide AI agents with the clarity needed to produce high-quality code.

Can I use Spectre commands standalone?

Yes. While the workflow follows a logical sequence, every command (like /spectre:scope or /spectre:clean) can be run independently as needed for your specific task.

How does /spectre:handoff help with long sessions?

It saves your current session state and progress into a snapshot, allowing you to clear your context window to keep Claude fast and accurate without losing project data.

How does Spectre reduce 'AI slop' or low-quality code?

Spectre eliminates ambiguity by forcing specificity up front. By generating scope documents and technical plans before coding starts, it prevents the AI from guessing your intentions.

What are canonical docs in Spectre?

These are persistent Markdown files generated in your repository (e.g., docs/tasks/) that serve as the source of truth for the AI's context during development.

Spectre Agentic Workflow Guide

Name: Spectre Agentic Workflow Guide
Author: Codename-Inc

byCodename-Inc

•

116

•

Productivity & Workflow

Implements a structured, step-by-step product development workflow to minimize ambiguity and maximize code quality for AI coding agents.

Spectre (Scope, Plan, Execute, Clean, Test, Rebase, Evaluate) is a comprehensive agentic framework designed to eliminate 'AI slop' by enforcing structured inputs and canonical documentation. By guiding both users and Claude through a rapid waterfall process, it ensures that requirements are crystalline, plans are deeply researched, and code is validated against original specs before merging. This skill is essential for developers building complex features who need superior context management and a reliable, repeatable process for high-stakes software engineering with AI.

Key Features

01Automated generation of canonical documentation (Scope, UX Specs, Technical Plans)

02Parallel subagent execution for rapid, multi-threaded implementation waves

03116 GitHub stars

04Risk-aware testing and automated, severity-based code reviews

05Structured S.P.E.C.T.R.E. workflow for end-to-end feature development

06Intelligent context management via session handoffs and memory consolidation

Use Cases

01Building complex software features from scratch with clear boundaries and architectural designs

02Maintaining codebase hygiene through automated cleaning, linting, and conflict-aware rebasing

03Scaling AI development through parallel task execution and structured subagent coordination

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add codename-inc/spectre spectre-guide

For use in Claude.ai and ChatGPT

Download Skill