The Agent Skills Directory

[PROMPT_INJECTION]: The skill provides instructions to read and analyze external content (source code and documentation files), which creates an indirect prompt injection surface. An attacker who can modify the source code or existing documentation could potentially embed instructions to influence the agent's behavior during the update process.
Ingestion points: The agent reads the documentation file at ${file} and analyzes source code at the component_path extracted from that file.
Boundary markers: Absent; there are no instructions to use delimiters or to ignore embedded natural language commands within the source data.
Capability inventory: The skill is limited to reading local files and updating markdown documentation. It does not have network access, shell execution capabilities, or the ability to run external binaries.
Sanitization: Absent; the content read from files is analyzed directly without validation, filtering, or escaping.
[NO_CODE]: The skill consists entirely of a markdown file with metadata and instructions; no executable scripts, binaries, or configuration files are present.

update-oo-component-documentation