Sparc Spec

This is phase one of the SPARC methodology, the requirements gathering step before you write any code. It walks you through extracting functional and non-functional requirements, writing Given/When/Then acceptance criteria, documenting constraints like performance and security limits, and mapping edge cases. Everything gets stored in memory using the claude-flow MCP server so later phases can reference it. The output is a structured spec document with sections for requirements, acceptance criteria, constraints, and integration points. If you're the type who jumps straight to coding and regrets it later, this forces you to think through what success looks like first. Works best when you actually have a complex feature that needs planning, not for trivial changes.

skills.sh installs: 733
GitHub stars: 81.4K
last updated: Aug 21, 2026

View on GitHub

Install

$ npx -y skills add ruvnet/ruflo --skill sparc-spec --agent claude-code

Installs into .claude/skills of the current project.

Files

SPARC Specification Phase

Run Phase 1 of the SPARC methodology: define what must be built and how success is measured.

When to use

When starting a new feature or project that needs structured requirements gathering before any code is written. This phase produces the foundational specification that all subsequent phases (Pseudocode, Architecture, Refinement, Completion) build upon.

Steps

Initialize phase tracking — call mcp__plugin_ruflo-core_ruflo__hooks_intelligence_trajectory-start with metadata { "phase": "specification", "feature": "$ARGUMENTS" }
Check for prior work — call mcp__plugin_ruflo-core_ruflo__memory_retrieve with namespace sparc-state and exact key current-phase-{feature-slug} to see if this workflow exists. If it does, retrieve existing artifacts. If not, initialize state at that exact key with phase 1, status: "active", and startedAt and updatedAt set to the current ISO timestamp.
Search for similar patterns — call mcp__plugin_ruflo-core_ruflo__neural_predict with the feature description to find relevant past specifications and learned patterns

SKILL.md on GitHub

Details

Repository: github.com/ruvnet/ruflo
First seen: Jun 3, 2026