crosspost
Pass
Audited by Gen Agent Trust Hub on Mar 30, 2026
Risk Level: SAFEPROMPT_INJECTIONDATA_EXFILTRATIONCOMMAND_EXECUTION
Full Analysis
- [PROMPT_INJECTION]: The skill processes user-provided source content to adapt it for different social platforms. This ingestion of untrusted data without defined boundary markers or sanitization creates a surface for indirect prompt injection.
- Ingestion points: User-provided "source content" (SKILL.md).
- Boundary markers: The instructions lack explicit delimiters or instructions for the agent to ignore embedded commands within the user data.
- Capability inventory: The skill utilizes network-capable functions (
requests.post) and references external tools (x-api) for posting. - Sanitization: No content validation or sanitization logic is specified.
- [DATA_EXFILTRATION]: The skill includes example implementation code that uses the
requestslibrary to transmit data to an external API endpoint (https://your-crosspost-service.example/api/posts). This demonstrates the functional capability for external network operations to non-whitelisted domains. - [COMMAND_EXECUTION]: The skill instructs the agent to utilize external tools and APIs, such as the
x-apiskill and platform-native REST APIs, to automate the publishing process.
Audit Metadata