The Agent Skills Directory

[COMMAND_EXECUTION]: The SKILL.md file contains a mandatory 'Memory Protocol' section that instructs the agent to execute a shell command: cat .claude/context/memory/learnings.md. This bypasses standard tool usage by prescribing a specific bash command for state management.
[PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection due to its core functionality and toolset.
Ingestion points: The skill ingests code snippets provided by users for review and operates on local files specified in the target input parameter.
Boundary markers: There are no explicit instructions or delimiters used to separate user-provided code from the agent's internal instructions.
Capability inventory: The skill has access to powerful tools including Bash, Write, Edit, and Grep, allowing for arbitrary command execution and file system modification.
Sanitization: The skill does not implement sanitization or validation of the code it reviews to prevent embedded instructions from influencing the agent's behavior.

vue-expert