devrel-webinar-program
Pass
Audited by Gen Agent Trust Hub on Nov 2, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to process untrusted data from multiple sources, which presents an indirect prompt injection attack surface.
- Ingestion points: The agent ingests data through user-provided webinar metrics (SKILL.md, 'How to read the numbers'), community-sourced questions from Discord or forums (references/formats-and-exemplars.md), and partner-supplied registrant lists (references/handoff-and-consent.md).
- Boundary markers: The instructions do not specify the use of delimiters or safety warnings to ensure the agent ignores potentially malicious instructions embedded within user-supplied inputs or community questions.
- Capability inventory: The skill instructs the agent to store project context in a local file (
devrel-context.md) and generates business-critical documents based on these inputs. - Sanitization: No mechanisms for sanitizing, filtering, or escaping external content are provided before the data is used in generation or stored in persistent memory.
- [EXTERNAL_DOWNLOADS]: The skill references and cites performance benchmarks from several well-known industry services, including Goldcast, ON24, Demio, Livestorm, and BrightTALK. These references are used for informational diagnostic purposes and do not involve the execution of external code.
Audit Metadata