The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses a local script scripts/doc_audit.py which executes git commands via subprocess.run to determine the project root and file modification times. These calls are implemented securely using argument lists and are standard for the skill's purpose.
[PROMPT_INJECTION]: Inherent surface for indirect prompt injection exists as subagents ingest project files and code.
Ingestion points: Search subagents read markdown files in the docs/ and manual/ directories and scan codebase exports as specified in references/agent-dispatch.md.
Boundary markers: Prompts for subagents do not currently include explicit boundary markers or instructions to disregard embedded commands in the data.
Capability inventory: The skill allows for file system modifications (writing/updating docs) and local shell execution for auditing.
Sanitization: The skill does not implement sanitization for the ingested documentation content before processing.

doc-maintenance