The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill instructs the agent to read and process data from external sources, specifically Dooray ticket logs and comments, which introduces a risk of indirect prompt injection.
Ingestion points: Dooray project tickets and logs referenced in references/dooray.md.
Boundary markers: The skill relies on MCP resources but does not provide specific instructions to the agent to disregard instructions found within ingested data.
Capability inventory: The skill includes a Python script (scripts/inspect_repository.py) that executes shell commands and provides instructions for database query evaluation.
Sanitization: The skill lacks explicit sanitization or filtering logic for external ticket content.
[DYNAMIC_EXECUTION]: The skill includes a local utility script scripts/inspect_repository.py designed to inspect Git repositories. This script uses the subprocess module to run commands at runtime.
Evidence: The git function in scripts/inspect_repository.py uses subprocess.run to execute shell commands.
[COMMAND_EXECUTION]: The provided repository inspection script executes system-level git commands. While restricted to the git binary, it accepts arguments that are executed in a shell context, representing a potential command execution vector.

daldaleng