The Agent Skills Directory

[PROMPT_INJECTION]: Indirect Prompt Injection Risk.
Ingestion points: The agent is instructed in SKILL.md to explore the codebase and investigate factual questions to reach a shared understanding.
Boundary markers: The instructions do not specify any delimiters or safety guidelines to ignore instructions that may be embedded within the processed codebase files.
Capability inventory: The skill involves reading project files and writing the final brainstorm brief to the local filesystem.
Sanitization: The skill does not mention any validation or sanitization of the content retrieved from the codebase before it is integrated into the agent's context.
[COMMAND_EXECUTION]: The skill instructs the agent to record output to dynamic filesystem paths using shell-like variable expansion, specifically ${TMPDIR:-/tmp}/lite-brainstorm-<slug>.md.

lite-brainstorm