The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes local project files to detect languages and missing workflows, creating a potential surface for indirect prompt injection.\n
Ingestion points: The skill reads the current working directory using Path.cwd() and inspects file structures with ls -la in SKILL.md.\n
Boundary markers: The instructions do not include specific delimiters or warnings to prevent the agent from interpreting instructions that might be embedded in the files it reads.\n
Capability inventory: The skill has Write permissions to create configuration files and uses Bash to execute inspection and validation commands.\n
Sanitization: Although the skill uses yaml.safe_load() for syntax validation, it does not sanitize the logical content of the data extracted from the project environment.\n- [DYNAMIC_EXECUTION]: The skill uses python3 -c to execute a validation check. The check utilizes yaml.safe_load(), which is a secure parsing method that prevents arbitrary code execution from malicious YAML content.\n- [COMMAND_EXECUTION]: The skill employs standard system and tool commands such as ls, gh workflow list, and pytest for validation and environment checks. These commands are appropriate for the skill's intended purpose and do not represent a security risk.

workflow-setup