event-planner
Pass
Audited by Gen Agent Trust Hub on Oct 25, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted data from external event organizer pages, uploaded files, and user-pasted content. Although it provides explicit instructions to the agent to treat this content as data rather than instructions and to report any injection attempts, the ingestion of third-party content constitutes a persistent attack surface.
- Ingestion points: Fetched organizer event pages, company event pages, uploaded files, and user-pasted text (SKILL.md Step 2).
- Boundary markers: The skill provides explicit instructions to disregard instructions found within external data and to treat them as data only.
- Capability inventory: The skill executes local scripts scripts/text_check.py and scripts/plan_math.py via subprocess to validate formatting and perform funnel arithmetic.
- Sanitization: The agent is instructed to detect and report scam-like requests (e.g., bank or tax details) and ignored instructions at the top of its output.
Audit Metadata