The Agent Skills Directory

[PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection by design as it executes tasks defined in external implementation plans.
Ingestion points: Loads plan files from the path specified in $ARGUMENTS or from the docs/plans/ directory.
Boundary markers: No specific delimiters or safety instructions are used to distinguish plan tasks from potential malicious injections within the files.
Capability inventory: The skill utilizes Bash, Write, Edit, and the gh CLI to implement and commit code.
Sanitization: No sanitization or validation of the plan content is performed before the agent acts on the instructions.
[COMMAND_EXECUTION]: Executes shell commands to manage the software development lifecycle.
Evidence: Utilizes git for branch creation and commits, npx tsc for type checking, and gh pr create for repository interaction.
[EXTERNAL_DOWNLOADS]: Interacts with well-known registries to execute development utilities.
Evidence: Executes npx and npm run commands, which may fetch packages from the official NPM registry during quality checks.

engineer-work