web-app-security
Pass
Audited by Gen Agent Trust Hub on Aug 27, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill employs a Python script to fetch and display HTTP response headers from arbitrary URLs. This creates a surface for tool output poisoning where a malicious server could return headers containing instructions designed to influence the AI agent's subsequent analysis or reporting.
- Ingestion points: The output of
scripts/inspect-security-headers.pyis read into the agent's context. - Boundary markers: The script provides clear prefixes for each header (e.g.,
content-security-policy: ...), but the skill instructions do not explicitly warn the model to ignore instructions found within the data. - Capability inventory: The skill is designed for auditing and reporting using the
assets/security-review.mdtemplate. It does not explicitly grant file-writing or further network capabilities beyond the inspection script. - Sanitization: The Python script prints header values verbatim without escaping or sanitizing potential instruction-like strings.
- [COMMAND_EXECUTION]: The skill includes
scripts/inspect-security-headers.py, which is a Python script that performs network requests usingurllib.request. While this is the intended functionality of the tool, it allows the agent to interact with external network resources.
Audit Metadata