tw-research-grounded-dialogue
Pass
Audited by Gen Agent Trust Hub on Jun 17, 2026
Risk Level: SAFEPROMPT_INJECTION
Full Analysis
- [PROMPT_INJECTION]: The skill exhibits a vulnerability surface for indirect prompt injection due to its iterative data ingestion model.
- Ingestion points: External data enters the agent context through the WebSearch tool as defined in SKILL.md Step 3.
- Boundary markers: The instructions lack explicit delimiters or mandatory safety prompts to ensure the agent ignores instructions potentially embedded in search results.
- Capability inventory: The skill is granted access to high-privilege tools including Bash and Write in the SKILL.md frontmatter. While used for legitimate archiving in Step 6, this increases the potential impact of an injection.
- Sanitization: There is no evidence of content validation or escaping of retrieved external data before it is interpolated into the dialogue templates.
Audit Metadata