The Agent Skills Directory

[COMMAND_EXECUTION]: The skill utilizes several shell-based tools including git (to verify remote URLs), gh (to search for repositories on GitHub), npm (to check package versions), and pnpm/bun (to execute package updates). These operations are restricted to the vendor's own accounts and are essential for the skill's primary purpose of cross-repo dependency management.
[INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted data from external sources such as GitHub search results (gh search code) and NPM registry metadata (npm view). This data is used to identify which repositories require updates and where they are located on the local filesystem. While this represents a standard attack surface for indirect prompt injection, the risk is mitigated by the skill's restricted scope (only repositories owned by remorses) and the fact that the agent is performing expected maintenance tasks.

bump-holocron-consumers