dashboard-expert

Pass

Audited by Gen Agent Trust Hub on Oct 12, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill ingests and processes content from Mixpanel dashboards, such as text card markdown and report metadata, which acts as a potential injection vector if the dashboard contains adversarial instructions.
  • Ingestion points: Dashboard content is retrieved via ws.get_dashboard() and parsed in SKILL.md and references/content-and-layout.md, specifically reading the markdown field from text cells.
  • Boundary markers: The skill does not instruct the agent to use specific delimiters or to disregard instructions found within the ingested dashboard content.
  • Capability inventory: The skill possesses extensive capabilities including code execution via Bash (Python and mp CLI), and the ability to Read, Write, and Edit local files.
  • Sanitization: There is no evidence of sanitization, filtering, or validation of the ingested strings from the dashboard before they are processed by the agent.
Audit Metadata
Risk Level
SAFE
Analyzed
Oct 12, 2026, 03:24 PM