The Agent Skills Directory

[COMMAND_EXECUTION]: The skill provides numerous shell one-liners and instructions for the agent to execute commands to audit the state of the project's documentation.
Uses tools like grep, ls, cat, git remote -v, awk, and sed to verify file metadata, frontmatter, and content.
Instructs the agent to execute a project-local script via bash scripts/board-snapshot.sh to update the session state.
References the execution of Node.js scripts such as npm run build:docs for generating API documentation.
[PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection because it processes and acts upon untrusted data from project files.
Ingestion points: The agent reads and parses content from Markdown files in docs/plans/ and the docs/BOARD.html file (SKILL.md).
Boundary markers: There are no explicit markers or instructions provided to the agent to treat the content of these files as untrusted or to ignore embedded instructions.
Capability inventory: The skill has access to the file system, shell execution (bash, git, npm), and describes automated publishing workflows in .github/workflows/docs.yml (SKILL.md).
Sanitization: The skill does not define any sanitization or validation logic for the external content it processes.

roblox-docs-and-writing