The Agent Skills Directory

[REMOTE_CODE_EXECUTION] (CRITICAL): The installation guide in SKILL.md instructs users to install just using a piped bash command: curl --proto '=https' --tlsv1.2 -sSf https://just.systems/install.sh | bash. Piped shell execution from remote sources is a high-risk pattern that can lead to arbitrary code execution if the source or connection is compromised.
[DATA_EXFILTRATION] (HIGH): The script scripts/shell_to_just.py accesses and reads the user's shell history files (~/.zsh_history, ~/.bash_history, ~/.history). These files frequently contain sensitive information, including API keys, passwords used in command-line arguments, and private directory structures, posing a significant data exposure risk.
[COMMAND_EXECUTION] (MEDIUM): The skill is designed to generate and execute Justfiles which have the capability to run arbitrary shell commands on the local system.
[INDIRECT_PROMPT_INJECTION] (LOW): The skill possesses a surface for indirect prompt injection by processing external data into executable scripts. 1. Ingestion points: Makefile, package.json, and shell history files processed by the scripts. 2. Boundary markers: Absent; there are no delimiters or instructions to ignore embedded commands in the source files. 3. Capability inventory: Subprocess execution of any recipe defined in the generated Justfile. 4. Sanitization: The scripts sanitize recipe names but do not validate or sanitize the shell commands themselves.

justfile