starling-watch

Pass

Audited by Gen Agent Trust Hub on Oct 9, 2026

Risk Level: SAFENO_CODEINDIRECT_PROMPT_INJECTION
Full Analysis
  • [SAFE]: No security risks were identified in the analyzed content. The skill operates as a set of strategic guidelines and reporting templates for a research persona.
  • [NO_CODE]: The skill is composed entirely of markdown documentation and contains no executable scripts, system-level commands, or binary files, which effectively eliminates common attack vectors such as remote code execution.
  • [INDIRECT_PROMPT_INJECTION]: The skill specifies a workflow for ingesting and analyzing untrusted content from public web sources (social media, blog posts, and community forums) during the 'SENSE' phase. This represents an attack surface for indirect prompt injection where external data could attempt to influence agent behavior. Ingestion points: Web search results retrieved via the agent's search tools (Phase 2: SENSE). Boundary markers: Absent; the instructions do not define specific delimiters for search result data. Capability inventory: No scripts or tools with file-writing, command execution, or outbound network capabilities (beyond the search tool itself) are present. Sanitization: Absent; no explicit validation of the retrieved content is defined.
Audit Metadata
Risk Level
SAFE
Analyzed
Oct 9, 2026, 01:31 AM