Flutter Security Audit: Find Leaks Before Attackers Do
Generic security tools don't know Flutter, and generic AI reviews call everything "critical".
Free
Works with the AI Agents you already use
Flutter Security Audit: Find Leaks Before Attackers Do
Example session with this skill installed
security audit this flutter app
- Read your context and instructions
- Compiled the flutter security audit
Summary
Real triage output from the bundled scanner on the intentionally vulnerable test app in the repo (tests/fixtures/vuln_app). The agent then validates each lead against the code and writes the final report with fixes.
Scanner leads (excerpt)
flutter-security-audit · vuln_app · 6 files · 25 rules
CRITICAL FSA-BE-001 Firebase rules open to everyone
firestore.rules:5
allow read, write: if true;
CRITICAL FSA-SEC-002 Supabase service_role key referenced in the client
lib/main.dart:15
const supabaseAdminRole = 'service_role';
CRITICAL FSA-NET-001 TLS certificate validation disabled
lib/main.dart:44
client.badCertificateCallback = (cert, host, port) => true;
HIGH FSA-AND-001 Android app marked debuggable
android/app/src/main/AndroidManifest.xml:5
HIGH FSA-STO-001 Sensitive value stored in SharedPreferences
lib/main.dart:22
await prefs.setString('access_token', token);
HIGH FSA-CRY-003 Hardcoded encryption key or IV
lib/main.dart:38
HIGH FSA-SEC-007 .env file bundled as a Flutter asset
pubspec.yaml:12
MEDIUM FSA-LOG-001 Sensitive value written to logs
lib/main.dart:23
print('login ok, token=$token');
MEDIUM FSA-BE-002 Client-side authorization flag
lib/main.dart:27
isPremium = true;
MEDIUM FSA-WEB-001 JavaScript bridge exposed to a WebView
lib/main.dart:49
MEDIUM FSA-WEB-002 WebView loads a URL from a variable
lib/main.dart:50
LOW FSA-BLD-001 Release build without code shrinking
android/app/build.gradle:4
What happens next
Leads are hints, not findings. The skill has the agent prove each one is reachable in a release build, drop false positives, map the rest to OWASP MASVS v2, and write REPORT.md plus findings.json with a code-level fix for every confirmed issue.
Connects securely to your tools. The creator never sees your data.
About this skill
Flutter apps leak in the same handful of ways, again and again: a Supabase service_role key or Telegram bot token shipped inside the APK, Firestore rules left in test mode, refresh tokens in SharedPreferences, badCertificateCallback => true left over from a dev server, or a deep link that reaches a WebView with a JavaScript bridge.
Generic security tools don't know Flutter, and generic AI reviews call everything "critical". This skill gives your agent a Flutter-specific playbook and makes it prove each finding before reporting it.
What it checks
- Secrets: hardcoded API keys, service-role keys, bot tokens, passwords
- Storage: tokens in SharedPreferences, Android backups, insecure local files
- Network: cleartext HTTP, disabled certificate validation, ATS and cleartext settings
- WebViews and deep links: JavaScript bridges, unvalidated link handling
- Backend: Firebase and Supabase rules, client-trusted flags like
isPremium - Crypto: weak randomness, hardcoded keys and IVs
- Platform and build: exported Android components, debug and release settings
How it works
- A fast, zero-dependency scanner (Node 18+) collects leads across your project.
- The agent validates each lead against real code paths and discards false positives.
- You get a report with severity, OWASP MASVS mapping, proof, and a fix in your own code style.
Open source
Created and maintained by the author. The free, MIT-licensed source, GitHub Action and issue tracker live at https://github.com/MBotirjonDev/flutter-audit-agent
Find more skills like this on the Claude Skills Marketplace, the Codex Skills Marketplace and the Gemini Skills Marketplace, or browse Claude Skills for Security.
Requires Node.js for scanning. Cannot verify server-side logic if backend source or rules are missing. Scope is limited to the provided repository and its platform configurations.
How to use
- 1
Connect your agent to the Agensi MCP
Copy the prompt and paste it into Claude, ChatGPT, Cursor or any agent that supports MCP. It sets up the connection for you. Manual setup
- 2
Sign in to Agensi
Your agent shows a sign-in link. Sign in and click Allow.
- 3
Ask for what you need
Your agent loads this skill and uses it on your task. It always asks before buying a skill.
Reviews
No reviews yet
Be one of the first to try it. Every listed skill goes through the trust checks below.
Security scanned
Checked by our 8-point scan before listing
Fresh listing
Recently published to Agensi
Free to use
No account required to browse
Trust & safety
Security scan
Security scan passed 19 days ago
- Free with an Agensi account
- Software
Node.js
Git