The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses Bash scripts (scripts/start-server.sh and scripts/stop-server.sh) to manage the lifecycle of a local companion server.
[DYNAMIC_EXECUTION]: The skill executes a bundled Node.js script (scripts/server.cjs) using the local node runtime to provide a web interface for visual mockups and user interactions.
[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted data from multiple sources which could be manipulated to influence the agent's behavior.
Ingestion points: Project documentation, recent git commits, and browser interaction logs stored in the events file within the session state directory.
Boundary markers: The skill instructions do not specify the use of delimiters or explicit 'ignore' instructions when reading external project files or interaction logs.
Capability inventory: The skill has capabilities to execute shell commands, run a Node.js server, write files to the local project directory, and commit changes to a git repository.
Sanitization: While the server script uses basic path normalization (path.basename) to prevent directory traversal when serving files, there is no semantic validation or filtering of the information read from files or logs before it enters the agent context.

brainstorming