The Agent Skills Directory

[SAFE]: The skill is designed for architectural and code research using the github_codebase_search tool. The author metadata identifies 'morph', which differs from the provided context 'ujang19', but this is a minor inconsistency with no identified security impact.
[PROMPT_INJECTION]: The skill interacts with untrusted external content, which creates a surface for indirect prompt injection.
Ingestion points: Content retrieved from external repositories via the github_codebase_search tool (SKILL.md).
Boundary markers: The instructions do not define specific delimiters or 'ignore' instructions for the retrieved content.
Capability inventory: The skill does not employ dangerous capabilities such as local command execution, file system modification, or network exfiltration.
Sanitization: No sanitization or validation logic is specified for the data ingested from external repositories.

code-research