The Agent Skills Directory

[REMOTE_CODE_EXECUTION]: The scripts/connect.py file contains an installation routine for the Higgsfield CLI that downloads a shell script from https://raw.githubusercontent.com/higgsfield-ai/cli/main/install.sh and pipes it directly into the shell (sh). Piped remote execution is a high-risk pattern that can be exploited if the source repository or the transit connection is compromised.
[COMMAND_EXECUTION]: The script scripts/connect.py uses subprocess.run with shell=True to execute the remote installation string. While the command itself is statically defined in the code, using the shell to execute piped commands increases the risk of unexpected behavior compared to list-based subprocess execution.
[EXTERNAL_DOWNLOADS]: The skill initiates downloads of external tools and configurations from non-whitelisted sources during its setup process, specifically for the Higgsfield AI and Codex CLI tools.
[INDIRECT_PROMPT_INJECTION]: The skill provides a surface for indirect prompt injection by taking user-provided prompts and passing them to external AI models and CLI tools without explicit sanitization or boundary markers. 1. Ingestion points: User-provided strings enter the skill via the --prompt and --text arguments in scripts/creator.py and scripts/media.py. 2. Boundary markers: Absent. User input is interpolated directly into prompts or command strings. 3. Capability inventory: The skill can execute shell commands (via ffmpeg, higgsfield, agy, codex) and perform network operations to various AI provider endpoints. 4. Sanitization: The skill does not perform content sanitization, relying instead on the safety filters of the external providers.

creator-character