Is it safe to scan my URL with LaunchGuide?

Scans target the public URL you submit. We do not claim to run penetration tests or exploit your systems. Findings are informational so you can fix issues with your AI coding agent.

Which agents are Live vs Coming?

url-security is Live today. repo-security, dependencies, auth, ai-surface, code-health, optimize, and standards-align are labeled Coming and will use the same findings → fix-prompt loop when they ship.

Which AI coding agents do fix prompts work with?

Fix prompts are written for Cursor, Claude Code, Antigravity, Codex, and similar agents. Paste the prompt into your IDE, apply the fix, then re-check.

What do I get for free vs paid?

A free scan shows severity counts. Starter ($49/mo, was $79) unlocks full URL findings and fix prompts, plus weekly URL monitoring where live. Pro ($99/mo, was $149) adds GitHub connect, weekly repo agents, inbox, and MCP as those features ship.

Do I need a GitHub repo to start?

No. Paste a live URL to scan today. Connecting a repo for weekly agents is a Pro feature marked Coming until it ships.

Where did planning / docs go?

Planning (clarify interview, engineering docs, and Cursor build workflow) remains available as a secondary product. Use the quiet Planning link in the nav, or visit /about.

Yearly totals are shown on the pricing cards ($470 Starter / $950 Pro). Annual billing is coming soon at checkout until Polar yearly SKUs are live.

The maintenance layer for AI-built apps.

Weekly audit agents. Findings with fix prompts for Cursor, Claude Code, Antigravity, Codex, and other AI coding agents.

Get started free →

Free scan shows severity counts · Upgrade for full findings + fix prompts

Cursor
Claude Code
Antigravity
Codex

Latest findings

Client LLM key patternCritical
Missing CSP headerWarning
Open redirect candidateWarning

Eight audit agents.
One maintenance layer.

Standing audits for AI-built apps. Live agents run today; Coming agents ship on the same findings → fix-prompt loop.

url-security

Live

Scan a live URL for security and AI-surface risks.

repo-security

Coming

Audit connected repos for security issues on a schedule.

dependencies

Coming

Flag vulnerable or outdated dependency patterns.

auth

Coming

Review auth flows, session handling, and access gaps.

ai-surface

Coming

Inspect LLM keys, prompt paths, and AI integration risks.

code-health

Coming

Surface maintainability and structural drift.

optimize

Coming

Highlight performance and efficiency opportunities.

standards-align

Coming

Check alignment with your project standards.

From URL to fixed
in one loop

Scan → agents → findings → fix prompt → re-check. Your coding agent writes the fix.

01
Paste a URL (or connect a repo)
Start with a live site scan today. GitHub-connected weekly agents are Coming on Pro.
02
Weekly agents run
Standing audit agents check security and AI-surface risks on a schedule — not a one-off checklist.
03
Review findings
See severity, scope, and context in one inbox. Free scans show counts; paid plans unlock full detail.
04
Paste a fix prompt → re-check
Generate a fix prompt for Cursor, Claude Code, Antigravity, Codex, or similar agents, then scan again.

Findings with fix prompts
for any AI coding agent

Generate a paste-ready prompt, drop it into your IDE, then re-check. No auto-PRs. No pentest claims.

Finding

Warningurl-security

Missing Content-Security-Policy header

Response headers on the scanned origin do not include a CSP. Add a restrictive policy to reduce XSS blast radius.

1. Generate fix prompt

Fix: add a CSP header on all HTML responses. Prefer nonce-based script policy. Report violations. Do not weaken existing security headers.

2. Paste into your agent

CursorPaste → re-check
Claude CodePaste → re-check
AntigravityPaste → re-check
CodexPaste → re-check

Starter and Pro. Scan free first.

Upgrade unlocks full findings and fix prompts. Annual billing coming soon at checkout.

Starter

$79$49/mo

Full URL findings, fix prompts, and weekly URL monitoring

Full URL findings + fix prompts
Weekly URL monitoring (Coming where not live)
Severity counts on free scan; upgrade unlocks detail
Email support

Get started free

Pro

$149$99/mo

Everything in Starter plus GitHub agents, inbox, and MCP

Everything in Starter
GitHub connect (Coming)
Weekly repo agents (Coming)
Findings inbox (Coming)
MCP access (Coming)

Get started free

Free scan shows severity counts; upgrade unlocks full findings and fix prompts. Subscribe in-app via Polar.

Common questions

Straight answers about URL scans, Live vs Coming agents, pricing, and where planning lives.

View pricing →

From the blog

Fresh guides and comparisons for founders building with AI coding tools.

View all posts →

Scan your app.
Fix it in your AI IDE.

Paste a URL to start a free scan, or create an account and upgrade when you need full findings and fix prompts.

Get started free →

Starter $49/mo · Pro $99/mo · Cancel anytime

The maintenance layer for AI-built apps.

Eight audit agents.
One maintenance layer.

url-security

repo-security

dependencies

auth

ai-surface

code-health

optimize

standards-align

From URL to fixed
in one loop

Paste a URL (or connect a repo)

Weekly agents run

Review findings

Paste a fix prompt → re-check

Findings with fix prompts
for any AI coding agent

Missing Content-Security-Policy header

Starter and Pro. Scan free first.

Starter

Pro

Common questions

From the blog

Top CodeGuide Alternatives for AI App Builders

How to Write a PRD for Lovable, Cursor, and Bolt

CodeGuide vs MakePRD: Complete Comparison

Scan your app.
Fix it in your AI IDE.

The maintenance layer for AI-built apps.

url-security

repo-security

dependencies

auth

ai-surface

code-health

optimize

standards-align

Paste a URL (or connect a repo)

Weekly agents run

Review findings

Paste a fix prompt → re-check

Missing Content-Security-Policy header

Starter

Pro

Top CodeGuide Alternatives for AI App Builders

How to Write a PRD for Lovable, Cursor, and Bolt

CodeGuide vs MakePRD: Complete Comparison

Scan your app.Fix it in your AI IDE.

Scan your app.
Fix it in your AI IDE.