System Scan
Automatically audit your codebase for security vulnerabilities, hardcoded secrets, and logical flaws.
The System Scan is Relia's core auditing engine. While the System Map helps you understand how your code works, the System Scan tells you what might be wrong with it. It performs deep static analysis, specifically tuned to catch common errors in AI-generated code.
Comprehensive Dashboard
The scan dashboard provides a high-level summary of your project's health, categorizing issues by type and severity.
System Scan Dashboard Placeholder (Showing Summary Counters: Critical, Warnings, Secrets, etc.)
Key Capabilities
Relia's scanner goes beyond simple linting by analyzing execution context and data flow.
1. Secrets & Sensitive Data Detection
Hardcoded API keys, passwords, and tokens are a major security risk, especially when AI tools lazily hardcode credentials during generation. Relia scans your entire project for exposed secrets.
Secrets Detection Placeholder (Showing list of detected API keys with file paths)
- Precise File Paths: We show exactly where the secret was found (e.g.,
src/gateway/test-helpers.server.ts:302). - Contextual Warnings: Issues are flagged as
WARNINGorCRITICALindicating if a fix is needed immediately.
2. Execution & State Inconsistencies
AI often hallucinates async logic or misses critical synchronizations. Relia detects these deep logical flaws.
Execution Vulnerabilities Placeholder (Showing WhatsApp Auto-Reply broken access control example)
- Affected Steps: The scanner traces the exact execution steps involved in the vulnerability.
- Exploit Scenarios: Relia explains how the flaw could be exploited (e.g., an attacker injecting a crafted value).
- Target Fix Location: We pinpoint the exact function or file that needs to be secured.
3. Output Propagation & Shape Mismatches
When components try to communicate, they must speak the same language. Relia checks if data shapes remain consistent across boundaries.
Output Propagation Issues Placeholder (Showing Truncated JSON payload error)
- Producer vs. Consumer Analysis: Relia compares what one function outputs against what the receiving function expects.
- Failure Prediction: We highlight what will happen when the mismatch occurs (e.g., a JSON parser failing to deserialize a truncated payload).
4. Branch Convergence Risks
When multiple logic paths merge, shared state needs to be managed carefully to avoid overwrites.
Branch Convergence Issues Placeholder (Showing Shared State File Overwrite Risk)
- Incoming Path Tracing: Relia visualizes all the different paths leading to a convergence point.
- State Overwrite Warnings: We detect if a shared resource (like an environment file) is being written to without proper append semantics, risking data loss.
Always address CRITICAL issues before deploying your application to production.
Next Step: Learn how to manage your audited projects in the Workspace.