false-green-tests
Pass
Audited by Gen Agent Trust Hub on Sep 18, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTIONREMOTE_CODE_EXECUTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill provides an installation command
npx @reticlehq/server@latest init. This command usesnpxto download and execute the initialization script for the Reticle server. As@reticlehqis the verified vendor of the skill, this is considered a standard and legitimate setup procedure for the associated tooling. - [INDIRECT_PROMPT_INJECTION]: The skill functions by ingesting and analyzing data from an active application environment, including DOM state, network requests, and framework internal state. This creates an ingestion surface for potentially untrusted data from the application under test. However, the skill is scoped to analysis and debugging, and no evidence of dangerous command execution or sensitive data exfiltration based on this input was found.
- [REMOTE_CODE_EXECUTION]: The documentation includes a
curlcommand to fetch text-based indexing fromhttps://docs.reticle.sh/llms.txt. This is a benign operation used to provide the agent with additional context from the vendor's official documentation site.
Audit Metadata