SpexFlow — AI skill for Claude Code
Node-based workflow for orchestrating codebase context and specifications for AI agents.
How to install SpexFlow
This entry records only its repository, not the path inside it, so there is no
exact command to give. Open shuxueshuxue/SpexFlow and copy the folder into
~/.claude/skills/, or the file into ~/.claude/agents/.
What SpexFlow does
Node-based workflow for orchestrating codebase context and specifications for AI agents.
Alternatives in AI
- Repomix — 📦 Repomix is a powerful tool that packs your entire repository into a single, AI-friendly file 22.7k ★
- Greptile — Code Search AI-powered codebase search and understanding 14k ★
- n8n Node Configuration — Node configuration with dependency rules and AI connections 3.6k ★
README
[!IMPORTANT] **SpexFlow is no longer maintained.** Its ideas continue — and go further — in **[SpexCode](https://github.com/shuxueshuxue/spexcode)**: where SpexFlow assembled context and a spec one feature at a time, SpexCode keeps a versioned spec tree inside your git repo and dispatches AI agents to build against it. Head there for active development.

[English](README.md) | [简体中文](README.zh.md)
SpexFlow is a visual context/spec workflow tool built on React Flow. It helps you turn a concrete feature request into:
- curated repo context (via code search or manual selection), then
- a high-quality implementation plan/spec from an LLM, then
- a clean prompt you can paste into a "fresh-context" code agent (Codex / Claude Code / etc.).
It's optimized for **"finish one well-defined feature in one shot"** rather than "keep everything in your head".
Screenshots
**A minimal workflow** (instruction → code-search → context → LLM):

**A larger canvas** where you keep reusable context blocks and rerun only the parts that changed:
**Two hops searching and planning**
**Spec dashboard** for managing spec runs and capturing outputs:

What You Build With It
SpexFlow loads a local code repo and lets you run a small node-based workflow:
- Instruction → produce/compose plain text input
- Code Search Conductor → generate multiple complementary search queries (one per downstream Code Search node)
- Code Search (Relace fast agentic search) → returns
{ explanation, files } - Manual Import → select files/folders and produce t
Related Skills
Generate Rules
An intelligent, zero-dependency Node.js CLI tool that automatically inspects any codebase and generates tailor
ClaudeCode Launchpad CLI
9 Windows & macOS 2-minute installer for Claude Code — auto-installs Node.js, Git & Claude Code; two-line live
Stacklit
One command gives AI agents instant codebase context. ~250 tokens replaces 50,000+ tokens of exploration. Auto
Promptscout
Enrich coding agent prompts with codebase context using a local LLM
Reporecall
Local codebase memory for Claude Code. Tree-sitter AST indexing (22 languages), hybrid keyword + vector search
Prompt Tower.Nvim
⚡ Turn your codebase into AI-ready context in seconds – the Neovim plugin that bridges the gap between your co
Related Agents
Opus Reviewer
Fresh-eyes review agent for the opus-build workflow Phase 4. Reviews one diff or branch against the codebase a
Dipeo Codegen Pipeline
Use this agent when working with DiPeO's complete code generation pipeline, including:\n- TypeScript model spe
Specifications Tracking Specialist
Specialist for maintaining comprehensive and accurate documentation in HISTORY.md and SPECIFICATIONS.md files.