linkedin-operator
Pass
Audited by Gen Agent Trust Hub on Aug 20, 2026
Risk Level: SAFEPROMPT_INJECTION
Full Analysis
- [PROMPT_INJECTION]: The skill reads external web data from LinkedIn profiles and messages, which creates a surface for indirect prompt injection where malicious content could influence the agent's actions.
- Ingestion points: LinkedIn UI state and page data are retrieved via the browser_capture tool as described in SKILL.md.
- Boundary markers: There are no instructions in SKILL.md or the reference files to use delimiters or ignore instructions embedded within the captured LinkedIn data.
- Capability inventory: The agent can execute browser actions (browser_act), navigate to new URLs (browser_navigate), and perform HTTP requests (http_request).
- Sanitization: The instructions do not specify any validation or sanitization of the data retrieved from the LinkedIn session before the agent acts upon it.
Audit Metadata