The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill ingests data from external sources including GitHub PR lists, security alerts, and repository manifest files which could potentially contain untrusted input. \n- Ingestion points: The skill reads output from gh pr list, gh api ... alerts, and gh pr diff as described in SKILL.md. \n- Boundary markers: No explicit delimiters or instructions are used to separate untrusted data from agent instructions. \n- Capability inventory: The skill utilizes capabilities for executing shell commands including git, gh, npm, pnpm, yarn, bun, cargo, go, and pip-compile. \n- Sanitization: No sanitization or validation of the fetched data is performed before it influences subsequent shell commands or file edits. \n- [COMMAND_EXECUTION]: The skill instructs the agent to execute various shell commands to manage Git branches, fetch GitHub repository data, and run package manager operations. \n- [DYNAMIC_EXECUTION]: The skill uses node -e to execute a runtime script that inspects package metadata within node_modules to verify dependency changes.

dependabot-batch