crosspost
Pass
Audited by Gen Agent Trust Hub on Apr 11, 2026
Risk Level: SAFEDATA_EXFILTRATIONPROMPT_INJECTION
Full Analysis
- [DATA_EXFILTRATION]: The skill provides a workflow for sending content to external social media APIs, documented in an example Python snippet using the requests library and a placeholder domain (your-crosspost-service.example).
- [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection because it processes untrusted user content and possesses the capability to write to external APIs. 1. Ingestion points: User-provided content for social media posts as described in the workflow in SKILL.md. 2. Boundary markers: Absent; there are no delimiters or instructions to ignore embedded commands. 3. Capability inventory: External network operations via requests.post as shown in the integration example. 4. Sanitization: Absent; no escaping or validation logic is defined for the processed content.
Audit Metadata