|
DevFlow
A meta-prompting, context engineering and spec-driven development system for Claude Code. Solves context rot.
|
82 % |
|
Thesis PM — The PM layer for spec-driven development
Thesis PM turns vague ideas into structured specs. Manage features, elaborate tasks and build with AI builder — Claude Code, Codex or Gemini.
|
82 % |
|
Speck - Opinionated Feature Specs for Claude Code
Transform spec-kit into a Bun-powered, Claude-native workflow with TypeScript-first tooling and slash command integration. Write better feature specs faster.
|
81 % |
|
claudeflow - AI-Powered Coding Workflows
Streamline ClaudeCode with prescribed prompt workflows. Build reliable AI coding pipelines via npm—open-source, extensible, and developer-first.
|
80 % |
|
RepoFrame: prepare repositories for AI coding agents
RepoFrame analyzes your repository locally, plans a task with Claude, and verifies the result with evidence: tests, lint, build and the git diff. Open source...
|
80 % |
|
SpecPad — IEC 62304 and FDA design-control evidence, in your git repo
Design-control evidence for medical device software, kept as schema-validated JSON in your git repo. Requirements, detailed design, risk, SOUP, cybersecurity...
|
79 % |
|
Claude Code Guide: features, commands, workflows, and tips
A bilingual React and TypeScript guide to Claude Code features, commands, workflows, MCP, hooks, skills, subagents, permissions, IDE integrations, CI, and pr...
|
79 % |
|
sysspec | sysspec
System specs first. AsyncAPI, OpenAPI, ODCS data contracts and Gherkin as enforceable system intent, with deterministic gates.
|
79 % |
|
specTamer — Cursor template that blocks coding until specs exist
specTamer is a Cursor template with OpenSpec-style workflow: proposal → specs → design/ADR → tasks → code. $12 one-time. OpenCode original stays free upstream.
|
78 % |
|
SpecFact - Review AI-assisted code before drift reaches PR or main
Open-source CLI for reviewing AI-assisted code against contracts, tests, and policy gates. Start with uvx in 60 seconds, then add IDE prompts, CI, and team w...
|
78 % |
|
tinyspec — Spec-driven development with Claude Code
A tiny framework for writing structured specifications that turn ideas into tracked implementation plans — powered by AI.
|
78 % |
|
Introduction - SpexCode
Spec-driven orchestration for your coding agents.
|
78 % |
|
Prompt Butter Jam — your prompt, marked up before your assistant sees it
Free tool that edits your prompt to any AI coding assistant: finds the vague goal, the missing constraint, the unstated assumption — and hands it back redlin...
|
78 % |
|
Vexa — a kanban board where Claude Code does the work
Desktop app for Windows that runs Claude Code agents through plan, code, tests, review, security and release — one task at a time, one git branch per task. U...
|
78 % |
|
Gencoded | AI Coding Assistant for Your Existing Architecture
Explore Gencoded, an AI coding assistant preview for architecture-aware changes, explicit contracts, unfiltered model choice, and precise review.
|
77 % |
|
Better Spec - Spec-Driven Development Toolkit
Better Spec is a spec-driven development toolkit that transforms your architectural vision into production-ready code. Create executable specifications that...
|
77 % |
|
Spec-driven handoff for Claude Code, Codex and Cursor | SpecWeave
Open-source CLI that keeps the spec, task ledger and handoff in git, so Codex or any agent continues when Claude Code hits its usage limit. Free, MIT.
|
77 % |
|
SpecFact Documentation
SpecFact Documentation - Keep backlog, specs, tests, and code in sync with automatic checks. For product owners, PMs, and developers. CLI reference, workflow...
|
77 % |
|
PromptRefiner — Better prompts, better AI answers
Chrome extension that turns rough ideas into clear, structured prompts in one click — right inside ChatGPT, Claude, and Gemini. Get the answer you wanted on...
|
77 % |
|
Guild — Spec-Driven Development for Claude Code
Guild installs spec-first workflows and specialized agent roles as .claude/ markdown files. Claude Code reads them natively. Design docs before code, zero in...
|
77 % |
Detected social media networks