The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses various shell tools including composer audit, bin/magento, find, grep, and curl to perform system-level audits and code scanning.
[DATA_EXFILTRATION]: The skill contains logic to probe for sensitive file paths such as .env, .git/config, and system.log. It performs network requests via curl to external domains to verify if these files are publicly accessible, which constitutes a network probing behavior.
[PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection due to its ingestion of untrusted external content without sufficient sanitization or boundary markers.
Ingestion points: Processes contents of local files in app/code/, output from binary execution (composer, magento), and headers from external network requests.
Boundary markers: No delimiters or instructions to ignore instructions embedded in the scanned data are provided.
Capability inventory: The skill has access to shell execution, filesystem manipulation, and database queries via MCP integrations.
Sanitization: There is no evidence of validation or escaping for data retrieved from the filesystem or network before it is processed by the agent.

security-scan