The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses a Node.js utility script, scripts/find-webnn-targets.mjs, to analyze the local workspace and identify potential integration targets. This script is limited to read-only operations on the project files and does not perform network access or execute external code.
[PROMPT_INJECTION]: An indirect prompt injection surface exists because the skill reads content from the user's codebase during the analysis phase. Ingestion points: File content is read by scripts/find-webnn-targets.mjs. Boundary markers: None are defined. Capability inventory: The agent can execute local scripts and perform filesystem modifications. Sanitization: No sanitization is performed on workspace content. This behavior is standard for development-focused skills and does not indicate malicious intent.
[SAFE]: The skill is consistent with its stated purpose. No evidence of credential harvesting, data exfiltration, or obfuscation was found during the analysis of the code and instructions.

webnn