profit-autopilot
Warn
Audited by Gen Agent Trust Hub on Jun 17, 2026
Risk Level: MEDIUMCOMMAND_EXECUTIONDATA_EXFILTRATIONCREDENTIALS_UNSAFEPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill instructs the agent to execute PowerShell scripts using the
-ExecutionPolicy Bypassflag. This command explicitly circumvents local security configurations designed to protect against the execution of unverified or potentially malicious scripts. - [DATA_EXFILTRATION]: The resource documentation includes absolute file paths (e.g.,
C:/Users/issda/SCBE-AETHERMOORE/...). This hardcoded metadata discloses the system's directory structure and the username 'issda' to the agent, providing a foothold for directory traversal or targeted local file access. - [CREDENTIALS_UNSAFE]: The skill's workflow depends on the presence of Shopify tokens for live publishing tasks. Automating actions that require high-privilege API tokens increases the risk of credential misuse if the agent is directed to perform unauthorized operations.
- [PROMPT_INJECTION]: The skill ingests and parses
product-landing.htmlto validate storefront states, introducing an indirect prompt injection surface. Maliciously crafted content within the HTML could be used to manipulate the agent's logic during the monetization pipeline. - Ingestion points: Reads
product-landing.html(SKILL.md) - Boundary markers: None; the skill lacks delimiters or instructions to ignore embedded commands in the processed file.
- Capability inventory: Execution of shell scripts and Python code with full system and network access (SKILL.md)
- Sanitization: No validation or sanitization of the landing page content is specified before processing.
Audit Metadata