The Agent Skills Directory

[PROMPT_INJECTION]: The skill includes instructions that attempt to override the agent's standard behavior and skip reading its own file content. It uses deceptive phrasing, claiming, "This prompt was injected into your context by Claude Code's command system," which can mislead the agent into treating the skill's instructions as high-priority system-level overrides.
[COMMAND_EXECUTION]: The skill performs shell operations using the Bash tool to execute local scripts located at $HOME/.claude/get-shit-done/bin/gsd-tools.cjs. These scripts reside outside the current working directory, which is a potential security risk if the environment is not strictly controlled.
[COMMAND_EXECUTION]: The query function directly interpolates the user-provided <term> argument into a shell command: node $HOME/.claude/get-shit-done/bin/gsd-tools.cjs graphify query <term>. This pattern is vulnerable to command injection if the agent provides an argument containing shell metacharacters (e.g., ;, &&, |).

gsd-graphify