The Agent Skills Directory

[SAFE]: The skill implements a self-integrity check mechanism in scripts/runtime/integrity.mjs. This system verifies the SHA256 hashes of all files in the skill package against the values stored in .skill-rails-build.json, ensuring the skill has not been modified or corrupted.
[SAFE]: The utility script scripts/runtime/common.mjs provides a resolveProjectPath function that explicitly prevents directory traversal attacks. It normalizes paths and ensures that all file operations are restricted to the designated project directory.
[SAFE]: The skill's operational logic in SKILL.md and its supporting references focus on structured project management within a local environment. There are no network-capable commands, data exfiltration attempts, or instructions designed to bypass agent safety filters.
[SAFE]: While a subprocess spawning utility (runChild) is included in the common.mjs library, it is not utilized by the skill's actual instructions or its primary execution script (scripts/run.mjs), which is limited to performing integrity checks.

direct