The Agent Skills Directory

[REMOTE_CODE_EXECUTION]: The skill bootstrap sequence includes instructions to download and execute a shell script from https://raw.githubusercontent.com/higgsfield-ai/cli/main/install.sh by piping it directly to sh. This allows for arbitrary code execution from a remote source during tool setup.
[EXTERNAL_DOWNLOADS]: The skill fetches executable installation scripts and configuration files from the higgsfield-ai GitHub organization, which is an external dependency.
[COMMAND_EXECUTION]: The skill makes extensive use of the Bash tool to interface with the higgsfield CLI and local utilities like mktemp, printf, and jq to manage data payloads and job submission.
[PROMPT_INJECTION]: The skill has an indirect prompt injection attack surface via the marketing-studio products fetch and brand-kits fetch features that process data from arbitrary external URLs. Ingestion points: Content from user-supplied URLs fetched in SKILL.md and references/marketing-products.md. Boundary markers: None identified in the provided instructions. Capability inventory: Execution of shell commands and network requests via the higgsfield CLI. Sanitization: No specific content sanitization or validation of retrieved website data is described.

higgsfield-generate