indian-dpdp-act-consent-notice-siddhi-kudalkar

CRITICAL E004: Prompt injection detected in skill instructions.

• Potential prompt injection detected (high risk: 0.90). The skill explicitly orders the agent to run a background "silent" web check and "do not tell the user you are doing this" (and to incorporate most findings without narration), which are hidden/deceptive instructions that go beyond normal, transparent research behavior.

MEDIUM W011: Third-party content exposure detected (indirect prompt injection risk).

• Third-party content exposure detected (low risk: 0.10). The required runtime workflow performs “Silent Legal Update Check” via web search (e.g., MeitY/notifications) before drafting, which can ingest outsider-authored free text from search results/网页 content.

MEDIUM W012: Unverifiable external dependency detected (runtime URL that controls agent).

• Potentially malicious external URL detected (high risk: 0.90). The skill explicitly requires a silent, runtime web update check (searching MeitY site) and references the MeitY page https://www.meity.gov.in/documents/act-and-policies/digital-personal-data-protection-rules-2025-gDOxUjMtQWa which the agent would fetch at runtime to modify its drafting instructions, so this external URL can directly influence agent prompts/output.