research-agent
Pass
Audited by Gen Agent Trust Hub on Apr 18, 2026
Risk Level: SAFENO_CODE
Full Analysis
- [NO_CODE]: The skill package contains only configuration files and markdown documentation (SKILL.md, workflow.yaml, automation.example.yaml). No executable scripts (.py, .js, .sh) or binaries are included, eliminating the risk of direct command execution or local persistence.
- [DATA_EXFILTRATION]: The skill defines tools for 'websearch' and 'webfetch', which are used to retrieve external data. This behavior is consistent with the stated purpose of a research agent and does not involve the transfer of sensitive local files or credentials.
- [INDIRECT_PROMPT_INJECTION]: The skill processes external data from the web, creating a potential surface for indirect prompt injection.
- Ingestion points: Content retrieved from external websites via the 'webfetch' tool as described in SKILL.md.
- Boundary markers: Absent; there are no explicit instructions or delimiters defined to prevent the agent from following instructions found in retrieved web content.
- Capability inventory: Access to 'websearch', 'webfetch', and 'text_summarize' tools as listed in SKILL.md.
- Sanitization: Absent; the workflow describes synthesizing findings but does not mention filtering or escaping content from external sources.
Audit Metadata