The Agent Skills Directory

[SAFE]: No security threats or malicious behaviors were detected. The skill is designed for documentation maintenance using the author's official resources.
[DATA_EXFILTRATION]: The skill retrieves information from the vendor's official GitHub repository (elevenlabs-dx). This data fetch is consistent with the skill's primary purpose of maintaining evergreen documentation.
[COMMAND_EXECUTION]: Standard development tools including Git and the GitHub CLI are used to manage pull requests and branches within the vendor's workspace. These actions are limited to repository management tasks.
[PROMPT_INJECTION]: The skill processes external changelog data, which represents a surface for indirect prompt injection. This risk is addressed by a mandatory verification workflow:
Ingestion points: Changelog data is downloaded to /tmp/changelog-${CHANGELOG_DATE}.md.
Boundary markers: None explicitly implemented.
Capability inventory: Subprocess calls for git and gh to modify documentation and manage pull requests.
Sanitization: Step 4 of the workflow requires the agent to verify all proposed edits against canonical API reference documentation before committing changes, preventing the unvalidated adoption of instructions from the changelog source.

update-skills-from-changelog