SPECTRE is a slash command based workflow for Claude Code designed to help you do ONE THING more, faster, and with higher quality. It covers the complete software development lifecycle - from scoping a feature, finalizing user flows, writing the technical design, generating tasks, executing the tasks, code review, validating the work, cleaning up and testing the work, and finally generating documentation as Skills your agent auto-loads when relevant.

SPECTRE helps you get higher quality and more consistent results from your coding agent, while they work autonomously for much longer, so 10-100x'ing your typical output feels easy and more importantly, repeatable. The workflow includes commands like /spectre:scope for feature scoping, /spectre:plan for implementation planning, /spectre:execute for parallel execution, /spectre:clean for code cleanup, /spectre:test for risk-aware testing, /spectre:rebase for safe merging, and /spectre:evaluate for architecture review and knowledge capture.

SPECTRE maintains session memory across sessions using the /spectre:handoff command, which generates status reports that get automatically loaded into your context window for the next session. The system uses specialized subagents including @spectre:dev for implementation, @spectre:analyst for understanding code, @spectre:finder for locating code, @spectre:patterns for finding reusable patterns, @spectre:web-research for web research, @spectre:tester for test automation, and @spectre:reviewer for independent code review.

The workflow generates canonical documents stored in docs/tasks/{topic/feature}/specs including scope.md for requirements, ux.md for user flows, plan.md for technical design, tasks.md for specific tasks, code_review.md for feedback, gaps.md for validation gaps, and .claude/skills/{feature_name}/skill.md for agent auto-referencing. SPECTRE's learning system uses SessionStart hooks to automatically load relevant skills based on trigger words, ensuring knowledge compounds across sessions instead of resetting.

SPECTRE is designed for product builders working on software development projects of any size, from brand new codebases to projects with hundreds of thousands of lines of code. It has been tested building websites, React Native apps, native desktop apps, and personal software, providing a single workflow that works on both small and large features without being overwhelmed with process.