The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses the Bash tool to execute repo-native commands for testing, build, and linting processes found within the project's configuration files (e.g., package.json, Makefile).
[COMMAND_EXECUTION]: The skill modifies project files using the Write tool to apply diagnostic-based fixes.
[PROMPT_INJECTION]: The skill has an indirect prompt injection attack surface:
Ingestion points: Reads repository configuration files, documentation, and verification command outputs from the local environment.
Boundary markers: Lacks explicit delimiters to differentiate between trusted instructions and untrusted data from repository content.
Capability inventory: Uses Read, Write, and Bash tools to execute commands and modify the file system.
Sanitization: No validation or sanitization is performed on repo-native commands or fix content before execution or writing.

ultraqa