pressreleasesonline

Pass

Audited by Gen Agent Trust Hub on May 7, 2026

Risk Level: SAFEDATA_EXFILTRATIONCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [DATA_EXFILTRATION]: The skill transmit s user-provi d e d website UR L s and email addresses to the externa l AP I at https://press releases.onlin e. This transmission is a functiona l req u i r e m e n t for generatin g and hostin g press releases via the service.\n- [COMMAND_EXECUTION]: The skill utilize s curl command s to communicate with the service AP I for submission and confirmation of press release s u b m i s s i o n s.\n- [PROMPT_INJECTION]: The skill retrieve s and processe s conten t from thir d-part y website s to draft press release s, introducin g a surfac e for indirec t promp t injection.\n
  • Ingestion point s: Website conten t from user-provi d e d UR L s in SKIL L.md.\n
  • Boundar y marker s: No boundar y marker s or 'ignor e' instruction s are presen t in the skill instruction s.\n
  • Capabilit y inventor y: The skill is limite d to AP I interaction s via curl.\n
  • Sanitization: No sanitization step s are define d in the skill instruction s; it relie s on backend processin g.
Audit Metadata
Risk Level
SAFE
Analyzed
May 7, 2026, 02:25 AM