Every Layer of
Application Security

From APK binary heuristics to repository SAST and dependency CVE scanning, Strata Security covers mobile binaries and source code repositories alike.

Looking for the product overview instead? See the Commercial Platform page.

Platform Capabilities

🔍

Static Analysis Engine

No runtime instrumentation. Strata Security analyses APK/IPA binaries and source repositories purely from static signals — manifests, bytecode, source files, and metadata.

Binary & SourceZero runtime
🦠

VirusTotal Integration

When VirusTotal is configured, each mobile scan looks up the file's SHA-256 hash through the VirusTotal API. Detections appear as findings in your report.

MalwareHash lookup
🔑

Secret & Credential Detection

Regex-backed scanning surfaces hardcoded API keys, OAuth tokens, private keys, and connection strings across up to 5,000 files per repository, before they reach production.

SecretsCredentialsTokens
📦

Dependency CVE Scanning

Package manifests are parsed (npm, PyPI, Maven, Gradle) and every dependency is checked against the OSV database for known CVEs — with severity, fix version, and CVE IDs.

CVEsOSVnpm · PyPI · Maven
🏢

Team Workspaces

Invite your security team, organise scans into projects, and share findings across your organisation with role-based access control.

TeamsProjectsRBAC
⚡

Results in <60 Seconds

Optimised pipeline processes most APKs and IPAs in under 30 seconds. Repository scans — including SAST and dependency lookups — typically complete in under 60 seconds.

FastAsync
🔒

Private & Secure

Files are streamed directly to the analysis engine. Content hashing prevents duplicate processing. Files and results are permanently deleted when your plan's retention period ends.

EncryptedSHA-256GDPR-ready

See our full security practices →

📊

Unified Risk Scoring

Every assessment, mobile or repository, produces a risk score from 0 to 100 and a risk level (Low to High for mobile scans, Low to Critical for repository scans) based on weighted findings, dependency severity, and secret confidence.

Score 0-100Weighted

APK Analysis — In Depth

Permission Intelligence

Strata Security checks every declared permission against a list of high-risk Android permissions. Each high-risk permission becomes a High severity finding mapped to OWASP Mobile Top 10 (2024) M6 (privacy) or M8 (misconfiguration), and dangerous permissions raise the overall risk score.

Special attention is paid to critical permissions like READ_CALL_LOG, SYSTEM_ALERT_WINDOW, and REQUEST_INSTALL_PACKAGES.

Dangerous permission detection with severity ratings
High-risk capabilities: accessibility service, overlay, package install, write settings
SMS, call log, camera and microphone access
Dangerous permission count feeds the overall risk score
Permission Analysis
READ_CALL_LOGHIGH
SYSTEM_ALERT_WINDOWHIGH
REQUEST_INSTALL_PACKAGESHIGH
RECORD_AUDIOHIGH
WRITE_EXTERNAL_STORAGEMEDIUM
READ_EXTERNAL_STORAGELOW
INTERNETINFO
ACCESS_NETWORK_STATEINFO

Manifest & DEX Analysis

Beyond permissions, Strata Security inspects the full AndroidManifest.xml for exported components, debug flags, backup settings, and cleartext traffic declarations.

DEX string scanning catches hardcoded API keys and other embedded secrets that manifest-only scanners miss, and the scan also flags plain http:// URLs and checks whether the code is obfuscated.

Exported Activity / Service / Provider detection
android:debuggable and android:allowBackup flags
Cleartext traffic (usesCleartextTraffic)
Hardcoded secrets in DEX strings
Plain http:// URLs and private IP addresses in bundled files
Manifest Flags
android:debuggabletrue ⚠
usesCleartextTraffictrue ⚠
android:allowBackuptrue ⚠
Exported activities3 found
Flagged strings:
http://192.168.1.100/api/v1
google_api_key: [REDACTED]

IPA Analysis — In Depth

App Transport Security & Info.plist

iOS security largely lives in two places: Info.plist and the App Transport Security configuration. Strata Security reads both and reports ATS exceptions and unsafe settings.

NSAllowsArbitraryLoads is raised as a High severity finding under OWASP M5 (Insecure Communication), and per-domain exceptions are listed in the scan results.

NSAllowsArbitraryLoads and domain exceptions
NSExceptionAllowsInsecureHTTPLoads per domain
22 privacy permission keys read from Info.plist
Entitlement warnings: get-task-allow, Network Extension, keychain sharing
ATS Configuration
NSAppTransportSecurity: {
  NSAllowsArbitraryLoads: true // ⚠ M5
  NSExceptionDomains: {
    "api.example.com": {
      NSExceptionAllowsInsecureHTTPLoads:
      true // listed
    }
  }
}
Arbitrary loadsInsecure domainOWASP M5

Repository Scanning — In Depth

Secret & Sensitive File Detection

Strata Security walks up to 5,000 files in a cloned or uploaded repository (skipping dependency and build folders) looking for hardcoded secrets: API keys, OAuth tokens, private key material, database connection strings, and more, using a curated set of high-signal regex rules.

Sensitive file patterns (private keys, .env, credentials.json, SSH keys) are detected regardless of content and flagged before they can be exploited post-leak.

AWS, GCP, Azure, and generic API key patterns
OAuth client secrets and GitHub, GitLab, Slack and Stripe tokens
RSA / EC private key material
Committed .env files, service account JSON, and keystore files
Confidence scoring (high / medium) to minimise noise
Secret Findings
aws_access_key_idCRITICAL
stripe_live_secret_keyCRITICAL
github_oauth_or_server_tokenHIGH
database_url_with_credentialsHIGH
generic_api_keyMEDIUM
⚠ Sensitive file detected
config/service-account.json

SAST & Code Vulnerability Analysis

Strata Security's static application security testing engine scans Python, JavaScript, TypeScript, Java, Kotlin, and Swift, flagging injection sinks, SSRF, unsafe deserialization, weak cryptography, and other exploitable patterns.

Each code finding is mapped to an OWASP Web Top 10 (2021) category (or, for AI findings, an OWASP LLM category) and a CWE identifier, giving developers the exact remediation context they need without leaving the report.

SQL injection, command injection, and SSRF
XSS and unsafe HTML rendering
Insecure deserialization and unsafe eval / exec
Weak cryptography (MD5, SHA-1, ECB mode)
OWASP Web Top 10 (2021) and CWE mapping
SAST Findings
SQL InjectionCRITICAL
api/OrderDao.java:84A3CWE-89
Command InjectionHIGH
utils/run.js:22A3CWE-78
Weak HashMEDIUM
auth/hash.py:11A2CWE-328

Dependency CVE Scanning

Strata Security parses every package manifest it finds — package.json, requirements.txt, pom.xml, Gradle build files, and lockfiles — then queries the OSV (Open Source Vulnerabilities) database for known CVEs in each pinned version.

Results include the CVE IDs, the advisory's severity, the version that fixes it when the advisory lists one, and links to the advisory, giving engineers the upgrade target without additional research.

npm, yarn and pnpm (package.json + lockfiles)
PyPI (requirements.txt, poetry.lock, Pipfile.lock)
Maven and Gradle (pom.xml, build.gradle)
CVE IDs, severity, and fix version when one is published
Mapped to OWASP A6 — Vulnerable & Outdated Components
Dependency CVEs
lodash@4.17.20HIGH
CVE-2021-23337npmfix → 4.17.21
requests@2.25.1MEDIUM
CVE-2023-32681PyPIfix → 2.31.0
log4j-core@2.14.1CRITICAL
CVE-2021-44228Mavenfix → 2.15.0
Sources parsed: package.json, requirements.txt, pom.xml

AI & LLM Integration Analysis

As applications connect directly to AI providers, the integration boundary itself becomes a real attack surface. Strata Security detects 8 of the 10 OWASP Top 10 for LLM Applications categories — prompt injection, sensitive data sent unredacted to third-party AI APIs, unsafe model loading, and unsanitized AI output flowing into dangerous sinks.

Every finding maps to its OWASP LLM Top 10 category and a CWE identifier, while three categories (excessive agency, vector/embedding access control, unbounded consumption) rely on low-confidence heuristics that need human review.

Prompt injection — untrusted data interpolated into a prompt
Sensitive data (PHI/PII, raw request objects) sent to OpenAI/Anthropic APIs unredacted
Unsafe ML model loading and unpinned model provenance
AI response content flowing into eval() / innerHTML unsanitized
Secrets embedded in system prompts, and AI calls with no token limit
AI/LLM Findings
Sensitive Data Exposure to AI APIHIGH
services/support_bot.py:41LLM02CWE-200
Prompt InjectionHIGH
api/chat.js:17LLM01CWE-1427
Unbounded ConsumptionMEDIUM
api/chat.js:23LLM10CWE-400
8 of 10 OWASP LLM Top 10 categories covered — see full breakdown

Reports Clients Actually Read

Strata Security generates structured, professional-grade security reports — ready to send directly to development teams or executive stakeholders.

📑

Executive Summary

Every PDF opens with a high-level risk overview — overall score, finding counts by severity, and the highest-priority issues. Designed for both developers and executive stakeholders.

🔬

Finding Evidence

Each finding includes the evidence that triggered it and its OWASP category; repository findings add the affected file and line and a CWE identifier for full traceability.

🛠

Remediation Guidance

Every finding ships with actionable fix instructions — tailored to the language and framework. No vague recommendations, just the specific change needed.

⬇️

Multiple Export Formats

Export as a polished PDF for delivery (Pro and Team), structured JSON for CI/CD integration, or SARIF 2.1.0 for repository scans. Repository reports include OWASP Web Top 10 and CWE mapping sections.

PDFJSONSARIF

Start Scanning for Free

1 scan per day, no credit card required. Upgrade when you need more.

Prefer to explore hands-on first? Try the Decision Center — free, no account required.