internal-comms
Pass
Audited by Gen Agent Trust Hub on Sep 14, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill provides instructions to ingest data from internal communication platforms and document stores, which exposes the agent to indirect prompt injection if those sources contain adversarial content.
- Ingestion points: Instructions in examples/3p-updates.md, examples/company-newsletter.md, and examples/faq-answers.md direct the agent to retrieve information from Slack, Email, Google Drive, and Calendar.
- Boundary markers: There are no instructions or delimiters provided to ensure the agent treats the retrieved data as untrusted or to ignore embedded instructions.
- Capability inventory: The agent is tasked with generating broad-reaching internal communications (newsletters, FAQs), which could be used to propagate malicious content if the injection is successful.
- Sanitization: No sanitization, filtering, or validation steps are included for the data retrieved from external tools.
- [SAFE]: No malicious command execution, data exfiltration to external domains, or obfuscation techniques were identified in the skill files.
Audit Metadata