internal-comms
Pass
Audited by Gen Agent Trust Hub on Sep 4, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill instructions direct the agent to ingest and summarize untrusted data from various external sources including Slack messages, Google Drive documents, and emails. This creates a surface for indirect prompt injection where an attacker could place hidden instructions (e.g., in Slack reactions, doc comments, or email bodies) to influence the tone, content, or direction of the generated reports.
- Ingestion points: Slack posts, Google Drive docs, and Email threads accessed via agent tools as described in
examples/3p-updates.md,examples/company-newsletter.md, andexamples/faq-answers.md. - Boundary markers: Absent. The instructions do not provide specific delimiters or warnings to ignore instructions embedded within the source data.
- Capability inventory: The skill primarily uses read-only tools (Slack, Google Drive, Email, Calendar) to generate text summaries; no automated code execution or external network writing tools are explicitly requested in the instructions, though the agent's environment might possess them.
- Sanitization: Absent. The instructions do not specify any validation or filtering of the content pulled from these external sources before interpolation into the communication templates.
Audit Metadata