The Agent Skills Directory

[REMOTE_CODE_EXECUTION]: The run.js script implements a dynamic execution engine that writes arbitrary JavaScript code to a temporary file (e.g., .temp-execution-*.js) and executes it using the Node.js require() function. While this is the intended functionality of the skill, it creates a sink for executing any logic the agent generates.
[COMMAND_EXECUTION]: The skill executes shell commands during setup and execution. The run.js script uses execSync to run npm install and Playwright browser installation commands. Additionally, SKILL.md provides patterns for executing node commands to perform server detection.
[EXTERNAL_DOWNLOADS]: The skill is configured to download external dependencies and binaries. The package.json file specifies the playwright dependency, and the setup script downloads the Chromium browser binary from official registries via npx playwright install.
[PROMPT_INJECTION]: The skill facilitates the ingestion of untrusted data from external websites, creating a surface for indirect prompt injection attacks.
Ingestion points: Data enters the agent's context through Playwright navigation (page.goto) and extraction helpers (extractTexts, extractTableData in lib/helpers.js).
Boundary markers: There are no explicit instructions or delimiters used to separate processed web content from the agent's system instructions.
Capability inventory: The skill possesses the ability to execute JavaScript, run shell commands via execSync (in run.js), perform file system writes (writing temporary execution files), and conduct network operations.
Sanitization: There is no evidence of sanitization or filtering of external content before it is processed by the agent.

playwright-skill