The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill aggregates output from subagents (investigator, builder, reviewer) and incorporates this data into local wiki reports within the llmwiki/ directory. This creates a surface where instructions embedded in the files or data processed by subagents could theoretically influence the agent's behavior during the report generation or subsequent wiki updates.
Ingestion points: Data results from cavecrew-investigator, cavecrew-builder, and cavecrew-reviewer subagents are processed in the main thread.
Boundary markers: The skill does not define explicit boundary markers or instructions to ignore embedded commands within the subagent outputs.
Capability inventory: The skill performs file system operations, specifically creating and appending to files in llmwiki/wiki/draft/cave/, llmwiki/wiki/index.md, and llmwiki/wiki/log.md.
Sanitization: There are no explicit instructions for sanitizing or escaping the data retrieved from subagents before writing it to the wiki files.

cavecrew