The Agent Skills Directory

[PROMPT_INJECTION]: The skill is vulnerable to indirect prompt injection via repository-specific configuration files.
Ingestion points: The skill reads untrusted data from AGENTS.md and CLAUDE.md to extract version requirements and build/install commands.
Boundary markers: There are no instructions provided to use boundary markers or delimiters when processing the content of these documentation files.
Capability inventory: The skill uses the Bash tool to execute extracted commands, which can include arbitrary shell operations.
Sanitization: No sanitization or validation of the extracted commands is performed before they are executed.
[COMMAND_EXECUTION]: The skill executes shell commands to verify the development environment, including node --version, pnpm --version, pnpm install, and pnpm build. When the --fix argument is used, it may also execute nvm install or npm install -g pnpm@latest to attempt auto-remediation of detected issues.
[EXTERNAL_DOWNLOADS]: If the --fix flag is enabled, the skill may download the pnpm package from the official NPM registry using npm install -g pnpm@latest.

repo-healthcheck-node