The Agent Skills Directory

[COMMAND_EXECUTION]: The skill instructs the agent to execute caveman opportunities list to retrieve diagnostic data for optimization.
[INDIRECT_PROMPT_INJECTION]: The skill processes potentially untrusted data from external sources.
Ingestion points: Reads output from the caveman CLI and parses historical context from files matching the pattern .caveman/proposals/*.md.
Boundary markers: The instructions explicitly state that historical proposal files should be treated as "untrusted historic context."
Capability inventory: The skill has the capability to execute shell commands (caveman CLI) and perform file system read/write operations.
Sanitization: There are no explicit instructions for sanitizing or escaping the content retrieved from the CLI or local files before processing.

caveman-optimize