The Agent Skills Directory

[COMMAND_EXECUTION]: The skill leverages shell commands to automate development workflows, including executing project build scripts (npm run build), Git operations (git branch, git commit, git push), and GitHub CLI actions (gh pr create, gh pr merge) for creating and managing audit-related pull requests.
[PROMPT_INJECTION]: The skill has an attack surface for indirect prompt injection due to its core functionality.
Ingestion points: The skill ingests and analyzes the entire project root (app_directory), which may include untrusted content within source files, comments, or documentation.
Boundary markers: There are no explicit delimiters or boundary instructions used to isolate project data from the instructions provided to the specialist sub-agents.
Capability inventory: Sub-agents are launched in auto mode, allowing them to autonomously modify project files and execute commands to remediate found issues.
Sanitization: No sanitization or filtering is performed on the project source code or data before it is processed by the specialist agents during the audit.

superteam