The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill ingests markdown files produced by external extractions, which may contain malicious instructions disguised as data.
Ingestion points: Pattern libraries are read from docs/ux-extracts/ and .jez/artifacts/ux-extracts/ (specifically pattern-library.md files).
Boundary markers: The instructions do not define delimiters or specific 'ignore' directives for the agent when processing these files.
Capability inventory: The skill has capabilities to read multiple files and write new documents to docs/ux-comparisons/.
Sanitization: No explicit sanitization, filtering, or validation of the ingested pattern data is required before it is interpolated into the final output.

ux-compare