research
Pass
Audited by Gen Agent Trust Hub on Oct 6, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill describes a workflow that ingests data from external web sources, which is a known vector for indirect prompt injection.
- Ingestion points: The 'Gather' step in SKILL.md explicitly instructs the agent to use 'WebSearch' and 'WebFetch' to collect information from external websites.
- Boundary markers: The instructions do not specify the use of delimiters, XML tags, or explicit directives to ignore potential instructions embedded within the retrieved external content.
- Capability inventory: The agent has the capability to read local project files, perform web searches, and 'deposit' findings into a knowledge base or project artifacts (SKILL.md).
- Sanitization: No instructions are provided for sanitizing, filtering, or validating external content before it is synthesized into recommendations or persisted to storage.
Audit Metadata