stitch-mcp-get-screen
Pass
Audited by Gen Agent Trust Hub on Sep 20, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
- [COMMAND_EXECUTION]: The skill instructs the agent to execute a bash script by interpolating a URL from the
get_screentool directly into the command string. Even with quotes, this represents a potential command injection surface if the source data is not strictly validated. - [INDIRECT_PROMPT_INJECTION]: (1) Ingestion points: The skill downloads external HTML from Stitch for conversion. (2) Boundary markers: No instructions are provided to the agent to treat this content as untrusted or to ignore instructions within it. (3) Capability inventory: The skill has access to
Bash,Read, andWritetools. (4) Sanitization: There is no evidence of sanitization or content validation for the downloaded assets. - [EXTERNAL_DOWNLOADS]: The skill fetches resources from trusted Google domains including
stitch.withgoogle.comandstorage.googleapis.comas part of its core functionality.
Audit Metadata