The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted data from external GitHub issues and comments using the gh issue view command. This content is then interpolated into task contracts for subagents. Although the skill mandates strict isolation through git worktrees and requires the orchestrator to review all changes before committing, the processing of external natural language content creates an inherent surface for indirect prompt injection attacks where malicious instructions could be embedded in issue descriptions or comments.- [COMMAND_EXECUTION]: The skill utilizes shell commands including git and gh (GitHub CLI) to perform repository management tasks such as creating worktrees, switching branches, staging files, and creating pull requests. These commands are necessary for the skill's primary purpose and are constrained by specific rules, such as prohibiting git add . and requiring explicit file allowlists.

arc-parallel-implement