Audit and fix AI-generated apps, right inside  Claude Code

Discover.Fix.Ship.

FORGE is a CLI + MCP server for Claude Code. 16 Opengrep rules plus two targeted LLM agents audit your repo, triage the noise, apply fixes, and rescan until clean. Built for the solo builder shipping Supabase apps from Lovable, Bolt, and v0.

install forge
$ pip install vibe2prod
$ vibe2prod setup

Then run /forge inside Claude Code. Or skip the CLI and use the web UI.

0
Deterministic Opengrep Rules
0
Targeted LLM Agents
0
MCP Tools In Claude Code
0–100
Production Readiness Score

One MCP install. One slash command. Shipped.

CodeQL and Snyk hand you a report. FORGE runs inside Claude Code, applies the fixes, and reruns the scan until it’s clean.

1

Install

One command inside Claude Code.

Pip install vibe2prod, run the setup wizard, and four tools land in every Claude Code session: forge_scan, forge_status, forge_config, and forge_health. No web UI needed. Your repo never leaves your machine.

2

Scan

Opengrep + 2 targeted LLM agents.

16 deterministic Opengrep rules run first (secrets, SQL injection, XSS, path traversal, SSRF, auth bypass, CORS, insecure crypto, N+1, silent exceptions, more) at zero LLM cost. Then a Codebase Analyst (Minimax M2.5) maps your architecture and a Security Auditor (Claude Haiku 4.5) reasons over findings for context and severity.

3

Fix with /forge

The slash skill closes the loop.

Run `/forge` in Claude Code and the skill walks the full cycle: scan, triage findings, update .forgeignore for false positives with structured v2 entries, fix real issues with Claude, then rescan to verify. Autonomous, auditable, and you stay in control of the diff.

Set up in 60 seconds

One pip install, one interactive wizard. Point FORGE at Claude Code and it registers the MCP server, installs the /forge and /forgeignore slash skills, and you’re scanning in a minute.

vibe2prod setup
$ vibe2prod setup
vibe2prod
FORGE Setup Wizard
Configure FORGE for local code auditing.
Your code never leaves your machine — only LLM API calls go to OpenRouter.
Step 1/6 — OpenRouter API Key (optional)
Enter your OpenRouter API key (press Enter to skip)
Get yours at: https://openrouter.ai/keys
OpenRouter API key: ••••••••••••••••••••
Step 2/6 — Vibe2Prod Dashboard (optional)
Enable dashboard sync? [Y/n]: y
Step 3/6 — Data Sharing
Share anonymized suppression data? [Y/n]: y
Step 4/6 — Claude Code Integration
Claude Code detected!
Register FORGE as MCP server + install skills? [Y/n]: y
Register for all projects (user) or just this one? user
⠋ Registering MCP server...
MCP server registered (user scope)
/forge skill installed
/forgeignore skill installed
Configuration
API Key:
sk-or-v1****
Dashboard:
Enabled
Data Sharing:
Enabled
Claude Code:
Integrated (user scope)
Config:
~/.vibe2prod/config.json
Setup complete!
Next: vibe2prod scan ./your-project

MCP server registers

Four tools, straight into Claude Code

forge_scan

Run the full discovery pipeline on a repo path. Returns findings + Production Readiness Score.

forge_status

Live progress of an in-flight scan: phase, active agents, cost, and time.

forge_config

Active vibe2prod URL, forgeignore sharing consent, installed version.

forge_health

Health check. Reports whether OPENROUTER_API_KEY is set and the MCP server is reachable.

Then in Claude Code

/forge drives the loop

Claude Code · my-supabase-app
/forge
I’ll run the FORGE audit flow on this repo: scan, triage, fix, and rescan until clean. Starting with forge_scan.
forge_scan(path='.')
Scanning /Users/you/my-supabase-app… ✓ 16 Opengrep rules executed (0.8s) ✓ Codebase Analyst mapped 142 files ✓ Security Auditor reviewed 18 findings Critical: 2 High: 4 Medium: 7 Low: 5 Readiness Score: 62 / 100
One critical finding stands out. A client can directly write to the profiles.role column from the browser:
src/api/profile.ts:47
→ flagged by Opengrep + Security Auditor
→ classic AI-generated Supabase mistake
Want me to patch it and add a server-side guard? (Y/n)
Step 01

Install

$ pip install vibe2prod

Ships the CLI, MCP server, and bundled skills.

Step 02

Run the wizard

$ vibe2prod setup

Registers the MCP server + installs /forge & /forgeignore.

Step 03

Use /forge

$ claude → /forge

Audit, triage, fix, and rescan. All inside Claude Code.

CI / scripted install?

Headless mode skips the TUI and takes your key on the command line. Useful for Docker images, dotfiles, and AI agents running the setup themselves.

headless
$ vibe2prod setup --no-interactive --api-key sk-or-…

Not using Claude Code? Use the web UI.

Same FORGE engine, no install required. Sign in, paste a GitHub URL, and the scan runs in an ephemeral sandbox. $15 signup credit, pay-per-scan after.

How It Works

01

Install the CLI

pip install vibe2prod brings down the CLI, the MCP server, and the bundled /forge + /forgeignore skills.

02

Run the setup wizard

`vibe2prod setup` detects Claude Code, registers the MCP tools, and copies the slash skills into ~/.claude/commands/.

03

Run /forge in your repo

The skill calls forge_scan (16 Opengrep rules plus two targeted LLM agents) and walks you through every finding.

04

Fix + rescan until clean

False positives land in .forgeignore as structured entries. Real issues get patched, then the scan reruns to verify.

Not Just Severity — Actionability

Every finding is classified by what you should actually do about it, calibrated to your project's stage and context.

Must Fix
Exploitable vulnerabilities — fix before shipping
Should Fix
Real issues — prioritize this sprint
Consider
Valid observations — address when convenient
Informational
Noted for awareness — no action needed

Why builders pick FORGE

CodeQL, Snyk, and Semgrep stop at detection and expect you to upload a repo. FORGE lives in your editor, explains what to do, then fixes the code for you.

CapabilityOther scannersFORGE
Deterministic vulnerability scan
Runs inside Claude Code (no upload)
LLM-reasoned severity + context
Auto-applies the fix via /forge
Structured false-positive suppression
Rescans until the repo is clean
Tuned for Supabase / Lovable / Bolt / v0

Pay per scan. Or don’t.

These prices are for the managed web UI, and honestly, paying here is basically buying me a coffee. The markup keeps the sandbox running. For the cheapest scans, grab an OpenRouter key and switch to BYOK. You pay OpenRouter directly at cost.

No key at all? FORGE still runs the 16 Opengrep rules + deterministic scoring for free, forever. You lose the two LLM passes but the static scan is yours.

$15 free balance on signup. No credit card required.

Small repo
~$2
<5k LOC
Medium repo
~$6
5k-50k LOC
Large repo
~$8
50k+ LOC

Add funds anytime. $5 minimum deposit.

Questions? Answers.