Spec-driven development
for AI agents
Define what to build, then let the agent build it right.
Vibeflow separates thinking from coding. You define specs with guardrails, and the AI agent implements following your project's real patterns.
How Vibeflow works
Three commands. From codebase analysis to audited implementation.
analyze Deep-scans your codebase and learns your patterns, conventions, and architecture.
gen-spec Generates a spec with binary Definition of Done, scope, anti-scope, and budget.
implement Implements from the spec following your patterns, within budget, respecting anti-scope.
Plus: discover, prompt-pack, audit, quick, teach, stats
See it in action
Watch Vibeflow challenge a vague idea into a tight spec.
discover discover round 2 gen-spec Use cases
Vibeflow adapts to different roles and scenarios.
New feature in an existing project
You need to add a payment flow. Instead of letting AI guess, you define scope (Stripe only, no PayPal), anti-scope (no subscription management), and a 5-check DoD. The agent implements exactly that.
analyze → gen-spec → implement → audit Quick bug fix with controlled scope
A date formatting bug in the dashboard. You use quick mode, it generates a focused spec in seconds. The agent fixes the bug, nothing else. Audit confirms the fix.
quick → implement → audit Defining requirements before handing off
You have a feature idea but need to clarify scope before the dev team starts. Discover walks you through the hard questions and produces a clear PRD. No code involved.
discover → hand off PRD to team Onboarding a new developer
New dev joins the team. Instead of reading 200 files, they run analyze. Vibeflow produces a comprehensive guide to the project: architecture, patterns, conventions, all with real code examples.
analyze → dev reads .vibeflow/ All commands
Core pipeline plus power tools for every workflow.
analyze Deep-scans your codebase and produces documentation: stack, architecture, patterns, conventions, all with real code examples.
Understand your codebasegen-spec Generates a technical specification with binary Definition of Done (3-7 pass/fail checks), scope, anti-scope, and risks.
Define what to buildimplement Implements a feature from its spec with guardrails. Follows your patterns, respects budget and anti-scope, runs tests, self-verifies DoD.
Build with guardrailsaudit Verifies implementation against the spec. Checks DoD, patterns, and tests. Produces a PASS, PARTIAL, or FAIL report.
Verify the resultdiscover Interactive dialogue that turns a vague idea into a clear PRD. Challenges assumptions, cuts scope, forces decisions.
Clarify the ideaprompt-pack Creates a self-contained prompt for a coding agent in a separate session. Embeds your real patterns and conventions.
Delegate to another agentquick Fast-track for small tasks (under 4 files). Skips discovery, generates spec directly. For bug fixes and quick changes.
Small tasks, fastChoose your edition
Same methodology. Adapted to your agent.
GitHub Copilot
Prompts, agents, skills, and instructions for Copilot Chat.
npx setup-vibeflow@latest --copilot Cursor
Rules and skills for Cursor Agent. Slash commands via /skill-name.
npx setup-vibeflow@latest --cursor Claude Code
Plugin with slash commands for Claude Code CLI.
pe-menezes/vibeflow-claude By default the installer adds installed files and the .vibeflow/ folder to .gitignore so they are not committed. Remove the "Vibeflow" block from .gitignore if you want to track them in git.
Start building with specs
Define what you want, let AI execute, verify the result.
Created by Pedro Menezes
Get in touch
Have a question, feedback, or want to contribute? Send a message.