The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses shell commands (PowerShell) and Git to inspect the local environment.
Executes Get-ChildItem with recursion to map project structure and locate configuration files.
Executes various git log and git rev-list commands to extract metrics, contributor lists, and activity patterns.
Includes an instruction for the agent to use its "best judgment to execute additional commands" based on output, which grants significant autonomy in local command execution.
[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted data from the repository being analyzed, creating a surface for indirect injection.
Ingestion points: Reads repository documentation (*.md), configuration files (*.json, *.yaml), and raw commit messages/metadata from the Git history.
Boundary markers: The instructions lack specific delimiters or "ignore embedded instructions" markers when processing these external data sources.
Capability inventory: The agent has access to the editFiles tool (file writing) and shell command execution (PowerShell/Git).
Sanitization: There are no instructions for sanitizing or escaping content retrieved from the repository before it is used to generate the final narrative summaries.

repo-story-time