The Agent Skills Directory

[SAFE]: The skill provides structured instructions for an agent to act as an exploration judge. Its operations are limited to reading project context and writing design documentation to a specified directory within the repository.
[PROMPT_INJECTION]: The skill has an indirect prompt injection surface because it processes untrusted data from repository files and external references, which could influence the generated design artifacts.
Ingestion points: The LOAD_CONTEXT step reads repository files, documentation, configurations, AGENTS.md, and external references.
Boundary markers: The skill does not explicitly define delimiters or instructions to ignore embedded prompts in the ingested data.
Capability inventory: The skill utilizes the WRITE_ARTIFACT step to create or update files in the <repository-root>/docs/designs/ directory.
Sanitization: There is no mention of sanitizing or validating the content gathered from external references before synthesis.

ts-explore