php-crypto-audit

Pass

Audited by Gen Agent Trust Hub on Mar 25, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill functions as a static analysis tool intended to identify cryptographic flaws in PHP code, such as weak hashing algorithms (MD5/SHA1), insecure encryption modes (ECB), and hardcoded keys. This is a legitimate security auditing use case.
  • [SAFE]: No network activity, external downloads, or remote code execution patterns were identified. The skill references a local file for severity ratings and writes its output to a local directory.
  • [SAFE]: No obfuscation or persistence mechanisms were found in the instructions.
  • [PROMPT_INJECTION]: Indirect Prompt Injection Surface: The skill is designed to ingest and process untrusted external data (PHP source code). This creates a potential vector where malicious instructions embedded in the code being audited could influence the agent's behavior during the audit process.
  • Ingestion points: PHP source code files from external projects (SKILL.md).
  • Boundary markers: Absent; there are no instructions to use specific delimiters or warnings to ignore embedded instructions in the source files.
  • Capability inventory: The skill utilizes file system read access for analysis and file system write access to save reports in {output_path}/vuln_audit/. No subprocess execution or network capabilities are present.
  • Sanitization: Absent; the instructions do not specify any validation or filtering of the source code content before processing.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 25, 2026, 02:33 AM
Security Audit — agent-trust-hub — php-crypto-audit