update-x-profile

Pass

Audited by Gen Agent Trust Hub on May 6, 2026

Risk Level: SAFEPROMPT_INJECTIONNO_CODE
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill ingests user-provided text to update external social media profiles, which represents an indirect prompt injection attack surface.\n
  • Ingestion points: User-supplied input for fields such as bio, display name, location, and website URL as described in the /x/profile endpoint documentation in SKILL.md.\n
  • Boundary markers: The skill explicitly instructs the agent to show a before/after diff and wait for explicit user approval before executing any changes, providing a critical human-in-the-loop check.\n
  • Capability inventory: The skill utilizes API calls (PATCH /x/profile, PUT /x/profile/avatar, PUT /x/profile/banner) to modify account metadata on a third-party platform.\n
  • Sanitization: The instructions include a requirement to validate that image URLs use HTTPS and that image formats are restricted to JPG or PNG.\n- [DATA_EXPOSURE_AND_EXFILTRATION]: The skill uses the XQUIK_API_KEY environment variable for authentication. This is consistent with standard secure practices for handling sensitive credentials in AI agent skills.
Audit Metadata
Risk Level
SAFE
Analyzed
May 6, 2026, 10:35 PM