orca-chat-visuals
Pass
Audited by Gen Agent Trust Hub on Oct 27, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill facilitates the generation of HTML content based on potentially untrusted data provided by the user, creating a vulnerability where malicious inputs could influence the visual rendering.
- Ingestion points: User-provided data used to create charts, diagrams, tables, and reports described in
SKILL.md. - Boundary markers: No explicit delimiters or instructions are provided to separate untrusted user data from the HTML structure.
- Capability inventory: The agent is instructed to write files to the
ORCA_CHAT_SCRATCH_DIRand use the::orca-visualcomponent to render them. - Sanitization: There are no instructions for the agent to sanitize, escape, or validate user-controlled data before incorporating it into the HTML/JavaScript code.
- [EXTERNAL_DOWNLOADS]: The skill permits generated HTML pages to load scripts and styles from well-known and established service providers.
- Evidence: The instructions allow resource loading from
cdn.jsdelivr.net,unpkg.com,cdnjs.cloudflare.com,esm.sh,fonts.googleapis.com, andfonts.gstatic.com.
Audit Metadata