The Agent Skills Directory

[Local Repository and Git Operations]: The skill uses local git commands to manage branches and push code for analysis. It implements proactive security measures by checking for sensitive file patterns (e.g., .env, .pem, .key, .credentials) using grep before staging changes. Additionally, it mandates explicit user approval before performing any git push operations, ensuring the user maintains oversight of code sent to remote repositories.
[Processing External Analysis Reports]: The workflow retrieves reports from external services to suggest code fixes, creating a surface for indirect influence on code generation.
Ingestion points: External analysis data enters the agent context through the aws_devops_agent__list_journal_records and aws_devops_agent__get_release_readiness_report tools.
Boundary markers: The instructions do not specify the use of delimiters or 'ignore' instructions for the ingested report content.
Capability inventory: The skill possesses the capability to perform git commit and git push to apply and distribute generated fixes.
Sanitization: There is no mention of explicit sanitization or schema validation for the external report content before processing.
Mitigation: To manage this, the skill requires mandatory user review and approval before any suggested fixes are applied or pushed to the repository.

analyzing-release-readiness