The Agent Skills Directory

[PROMPT_INJECTION]: The skill contains instructions that use imperative language ("CRITICAL: Never use...") to override the agent's default behavior regarding tool usage for specific file paths. While this is intended as a safety measure to prevent secrets from entering the conversation context, instructions designed to mandate or bypass standard tool procedures are analyzed for potential behavior override patterns.
[INDIRECT_PROMPT_INJECTION]: The skill creates a surface for indirect prompt injection by taking user-supplied strings and placing them directly into shell commands.
Ingestion points: The agent is instructed to ask the user for a name and email in Step 3.
Boundary markers: No delimiters or safety instructions are used when interpolating these values into the shell command.
Capability inventory: The skill utilizes Bash capabilities to execute git config and glab commands.
Sanitization: There is no evidence of sanitization or validation for the user-provided strings before they are executed in the shell.
[COMMAND_EXECUTION]: The skill performs shell command execution to configure glab (GitLab CLI) and git. These operations are consistent with the skill's stated purpose of setting up a development environment.

cyrus-setup-gitlab