The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted Android source code, XML layouts, and resource files which serves as an ingestion point for potentially malicious instructions embedded in data.
Ingestion points: Affected screen code, XML/Compose code, adapter/list item code, and theme/style resources provided as inputs in SKILL.md.
Boundary markers: The skill does not define explicit delimiters or instructions to ignore embedded commands within the processed files.
Capability inventory: The skill utilizes codegraph explore, rg (ripgrep), and find for file system exploration and code analysis.
Sanitization: No specific sanitization or validation logic is defined for the content of the ingested files.
[COMMAND_EXECUTION]: The skill instructs the agent to use command-line utilities to search and explore the project structure.
Evidence: The workflow includes specific commands such as codegraph explore "<Composable, View class, or adapter>" and suggests fallback usage of rg and find in SKILL.md.

android-ui-regression-checker