The Agent Skills Directory

[REMOTE_CODE_EXECUTION]: The skill's setup instructions direct the agent or user to execute curl -fsSL https://cli.inference.sh | sh. This pattern pipes remote content directly into a shell, allowing for unverified code execution. While this is a vendor-provided installer, the method bypasses static review and integrity checks, posing a risk if the remote source or delivery path is compromised.\n- [EXTERNAL_DOWNLOADS]: The skill requires downloading external CLI binaries and configuration files from cli.inference.sh and dist.inference.sh to function.\n- [COMMAND_EXECUTION]: The skill uses the Bash tool to run the infsh CLI, which executes external applications for web searching (tavily, exa) and image generation (falai/flux-dev-lora).\n- [PROMPT_INJECTION]: The skill processes data from external search providers (Step 1) to build personas. This ingestion of untrusted data lacks boundary markers or sanitization, creating a surface for indirect prompt injection where malicious content in search results could influence the agent's output or subsequent tool use. Capability inventory includes Bash tool access via the infsh command.

customer-persona