cookie-attacks

Pass

Audited by Gen Agent Trust Hub on Sep 22, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest session cookie data from external, potentially untrusted web servers during auditing procedures, which represents an indirect prompt injection surface.
  • Ingestion points: Response headers from external targets (e.g., https://TARGET/login) are parsed to extract session tokens.
  • Boundary markers: The instructions do not define explicit boundary markers or "ignore instructions" tags for the data being processed.
  • Capability inventory: The skill uses curl and the Python requests library to replay these extracted tokens in subsequent network requests.
  • Sanitization: The skill uses a regular expression (grep -oP 'SessionID=[^;]+') to extract tokens, which provides basic structural filtering, though no formal sanitization is performed on the cookie values before they are used in headers.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 22, 2026, 05:34 PM
Security Audit — agent-trust-hub — cookie-attacks