The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted local code to provide architectural recommendations, which could be exploited via indirect prompt injection.
Ingestion points: The skill uses Glob, Grep, and Read to ingest code implementation details from the user's filesystem (specifically src/**/<name>*.ts) as part of its investigation step in SKILL.md.
Boundary markers: The skill does not employ specific delimiters or instructions to ignore embedded natural language commands within the files it analyzes.
Capability inventory: The skill possesses filesystem search and read capabilities (Glob, Grep, Read).
Sanitization: There is no evidence of content sanitization or validation before the ingested data is used by the agent to formulate responses.

angular-architecture