The Agent Skills Directory

[COMMAND_EXECUTION]: The skill relies on shell commands like git, grep, printf, and date to analyze corpus changes and update the knowledge base. These commands are legitimate for the skill's purpose and are scoped to the local corpus environment.
[PROMPT_INJECTION]: The skill possesses an indirect prompt injection surface. Ingestion points: Data enters the agent context through git diff output and local knowledge base files (.kb/summaries). Boundary markers: The workflow does not specify delimiters or instructions to ignore malicious content within the analyzed source files. Capability inventory: The agent can execute shell commands and modify local files. Sanitization: There is no mention of sanitizing or validating external content extracted from the corpus before it is used to update summaries.

kb-refresh