The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill utilizes npx @google/clasp to download and run the Command Line Apps Script Projects (clasp) utility. This targets the official package published by Google on the NPM registry.
[COMMAND_EXECUTION]: The skill instructs the agent to execute shell commands using the clasp CLI for project management tasks, including authentication, script creation, code synchronization, and web application deployment.
[DATA_EXFILTRATION]: Instructions explicitly include a "Red Line" policy advising users to avoid writing PII (names, IDs, phone numbers) into spreadsheets or script code, mitigating the risk of accidental data exposure via public web app URLs.
[PRIVILEGE_ESCALATION]: The skill addresses Windows execution policy restrictions by recommending npx to bypass global blocks or using Set-ExecutionPolicy -Scope Process -ExecutionPolicy RemoteSigned to limit script execution permissions to the current session only.
[INDIRECT_PROMPT_INJECTION]: The skill establishes an interface between the agent and external Google Apps Script projects.
Ingestion points: The agent reads script content and spreadsheet metadata from Google's servers.
Boundary markers: The skill instructs the agent to clarify user intent before creating scripts or overriding existing .clasp.json configurations.
Capability inventory: The agent can write to the local filesystem and push code to the user's Google account using authenticated tools.
Sanitization: The skill emphasizes using the user's primary language and confirming operations before execution.

clasp-setup