The Agent Skills Directory

[COMMAND_EXECUTION]: The skill executes multiple shell commands to manage the development environment, including kill-port to clear ports, starting Google Chrome with specific debugging and performance flags, and using git commands for diffing and status monitoring.
[EXTERNAL_DOWNLOADS]: The instructions direct the agent to install the @playwright/cli package globally from the official NPM registry if it is not already present in the environment.
[DYNAMIC_EXECUTION]: The skill utilizes a local debugger bridge to evaluate JavaScript expressions within the running application's context via the /debug/evaluate endpoint to harvest narration data.
[INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted data from git diffs and PR content to determine where to set breakpoints and what to narrate. This represents a potential surface where malicious code comments could attempt to influence the agent's narration.
Ingestion points: Git diff outputs (git diff, gh pr diff) and source code files.
Boundary markers: None specified in the prompt instructions.
Capability inventory: Shell command execution (curl, playwright-cli, git), file system modifications (writing tour artifacts and updating .vscode/settings.json).
Sanitization: The skill does not perform specific sanitization or escaping on the ingested diff content before processing.

pr-tour