stitch-mcp-list-screens
Pass
Audited by Gen Agent Trust Hub on Sep 20, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill ingests data from external tool outputs to facilitate subsequent actions.
- Ingestion points: The agent reads the
screensarray (includingnameandtitlefields) from thelist_screenstool output as described inSKILL.md. - Boundary markers: There are no explicit instructions for the agent to use delimiters or to disregard instructions potentially embedded within the retrieved screen metadata.
- Capability inventory: The skill uses extracted data (IDs) to execute further tools such as
stitch-mcp-get-screen. - Sanitization: The skill lacks requirements for sanitizing or validating string fields retrieved from the Stitch API before they influence the agent's logic flow.
Audit Metadata