image-fetcher
Pass
Audited by Gen Agent Trust Hub on Sep 23, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill navigates to external URLs and interacts with web content, creating a surface for instructions embedded in third-party websites to influence agent behavior.
- Ingestion points: External web content accessed through the Playwright MCP server during navigation and element selection.
- Boundary markers: The skill does not define clear boundaries or specific instructions to the agent to ignore content found within the visual assets or targeted sites.
- Capability inventory: Includes browser navigation, viewport manipulation, screenshot capture, and the ability to write files to specific reference/output directories.
- Sanitization: The skill proactively implements safety measures by instructing the agent to block the
file://protocol and private IP ranges (localhost, 127.0.0.1, 10.x, 192.168.x), which mitigates local file exposure and SSRF risks.
Audit Metadata