internal-comms
Pass
Audited by Gen Agent Trust Hub on Jul 10, 2026
Risk Level: SAFEPROMPT_INJECTION
Full Analysis
- [PROMPT_INJECTION]: The skill is vulnerable to indirect prompt injection because it is designed to ingest and process data from multi-user platforms like Slack, Email, and Google Drive. An attacker could place malicious instructions in these shared environments that the agent might inadvertently follow while generating communications.
- Ingestion points: Slack posts, Google Drive documents, Emails, and Calendar events referenced in 'examples/3p-updates.md', 'examples/company-newsletter.md', and 'examples/faq-answers.md'.
- Boundary markers: Absent. The instructions do not define delimiters to separate untrusted data from processing instructions.
- Capability inventory: The skill is capable of reading internal data and formatting it into public-facing communications (newsletters, FAQs).
- Sanitization: Absent. There are no filters or validation steps to ensure the gathered content does not contain instructions.
Audit Metadata