The Agent Skills Directory

[COMMAND_EXECUTION]: The skill executes shell commands via git and the GitHub CLI (gh) to automate repository management tasks including branch creation, commits, and pull request generation. These operations are core to the stated purpose of the Plot workflow.
[PROMPT_INJECTION]: The skill processes user input from $ARGUMENTS into shell commands, which presents an indirect prompt injection surface. The author mitigates this risk by requiring strict regex validation for the slug parameter.
Ingestion points: User-provided $ARGUMENTS in SKILL.md (Step 1).
Boundary markers: Explicit instructions require that the slug must match the regex [a-z0-9-]+.
Capability inventory: File system operations (mkdir, ln, writing markdown files) and subprocess execution (git, gh) are present in SKILL.md (Steps 3-7).
Sanitization: Validation is enforced for the slug parameter to prevent path traversal or shell injection in filenames and branches; however, the title is interpolated into git commit and gh pr create commands without additional specific sanitization instructions.

plot-idea