The Agent Skills Directory

[COMMAND_EXECUTION]: The skill executes various shell commands for project auditing, file management, and error recovery.
Evidence: Uses git status, git log, git branch, git mv, and git restore for standard operations and auditing.
Evidence: Rollback procedures in SKILL.md include high-impact commands such as git reset --hard and git push --force.
Evidence: Performs file system modifications and deletions using git mv and rm.
[PROMPT_INJECTION]: The skill possesses an indirect prompt injection surface as it analyzes untrusted data from the project directory.
Ingestion points: Specialist agents (clutter-analyst, nomenclature-enforcer, etc.) read project file names and contents to determine organization logic.
Boundary markers: The instructions do not specify the use of delimiters or 'ignore' instructions for the content of the files being analyzed.
Capability inventory: The skill has broad file-system modification privileges, including moving, renaming, and deleting files via the decision-integrator agent.
Sanitization: There is no mention of sanitizing file content or validating metadata before it influences agent decisions.
[SAFE]: Employs secure coding practices for configuration handling.
Evidence: references/archival-compliance-check.md explicitly mandates the use of yaml.safe_load() and prohibits the use of the unsafe yaml.load() method.

archive-workflow