The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill is designed to read and follow instructions from AGENTS.md and other repository source files. This creates a surface where repository content could theoretically influence the agent's behavior during the wiki update process.
Ingestion points: AGENTS.md, existing llm-wiki/ content, and repository source files (referenced in SKILL.md).
Boundary markers: The skill instructions suggest preserving boundaries between raw source and synthesis, but do not define explicit delimiters for instruction-following.
Capability inventory: The agent is authorized to write to llm-wiki/** and execute git diff --check -- llm-wiki (referenced in SKILL.md).
Sanitization: No specific sanitization methods are listed for content extracted from repository files, though the scope is restricted to the wiki directory.
[COMMAND_EXECUTION]: The skill utilizes git diff --check -- llm-wiki to validate changes before reporting. This is a standard validation command that operates on a specific directory, presenting minimal risk in this context.

llm-wiki-update