The Agent Skills Directory

[REMOTE_CODE_EXECUTION]: The skill instructs the agent to install the lpm tool using a piped shell script pattern.
Evidence: curl -fsSL https://raw.githubusercontent.com/gug007/lpm/main/install.sh | bash found in README.md and SKILL.md.
The resource originates from the skill author's repository and is required for the tool's functionality.
[COMMAND_EXECUTION]: The skill utilizes shell commands to manage project configurations and verify the environment.
Evidence: Commands such as command -v lpm, ls ~/.lpm/projects/*.yml, mkdir -p ~/.lpm/projects, and rm ~/.lpm/projects/<name>.yml are used for project lifecycle management.
Evidence: A shell loop using awk and sed is used to extract project root paths from configuration files.
[EXTERNAL_DOWNLOADS]: The skill fetches an installation script from an external source.
Evidence: https://raw.githubusercontent.com/gug007/lpm/main/install.sh.
[INDIRECT_PROMPT_INJECTION]: The skill processes project configuration data which could theoretically contain malicious instructions.
Ingestion points: The skill reads existing configuration files at ~/.lpm/projects/*.yml and accepts user input to modify them.
Boundary markers: The skill does not define specific delimiters for separating processed file content from instructions.
Capability inventory: The skill can execute shell commands (curl, ls, rm, awk, sed) and perform file system writes.
Sanitization: The SKILL.md file includes a 'Validation' section requiring the agent to verify fields such as port ranges, path existence, and naming conventions before writing configurations.

lpm-config