internal-comms

Pass

Audited by Gen Agent Trust Hub on May 4, 2026

Risk Level: SAFEPROMPT_INJECTIONNO_CODE
Full Analysis
  • [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection because it is designed to ingest and process data from external, potentially untrusted sources like Slack messages, emails, and shared documents.\n
  • Ingestion points: Content is pulled from Slack channels, email threads, and Google Drive documents via tool calls mentioned in examples/3p-updates.md, examples/company-newsletter.md, and examples/faq-answers.md.\n
  • Boundary markers: The instructions do not include specific delimiters or "ignore" instructions to prevent the agent from obeying commands found within the retrieved data.\n
  • Capability inventory: The skill uses these tools to summarize and format text; while it does not include dangerous shell commands, the agent's logic could be influenced by external input.\n
  • Sanitization: There is no explicit logic to sanitize or filter out instructional phrases from the source data.\n- [NO_CODE]: The skill consists entirely of Markdown instructions and configuration files; it does not include any executable scripts, binaries, or automated code generation.
Audit Metadata
Risk Level
SAFE
Analyzed
May 4, 2026, 02:53 AM
Security Audit — agent-trust-hub — internal-comms