gpc-user-management

Pass

Audited by Gen Agent Trust Hub on Mar 12, 2026

Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTIONEXTERNAL_DOWNLOADSCREDENTIALS_UNSAFE
Full Analysis
  • [COMMAND_EXECUTION]: The skill relies on shell command execution to interface with the gpc CLI tool. This is performed via the agent's command environment and a Node.js detection script (scripts/detect_gpc.mjs) that uses execSync to verify installation and configuration.
  • [PROMPT_INJECTION]: The skill's functionality for importing testers from CSV files (gpc testers import --file testers.csv) constitutes an indirect prompt injection surface by ingesting external data.
  • Ingestion points: Local CSV files containing tester email addresses.
  • Boundary markers: No explicit delimiter or instruction-ignore markers are specified for the CSV content.
  • Capability inventory: The agent can add, remove, and list users and testers in a Google Play developer account.
  • Sanitization: No sanitization or validation logic is present in the provided skill files.
  • [EXTERNAL_DOWNLOADS]: The skill suggests using npx to download and execute the @gpc-cli/cli package from the NPM registry if it is not already available in the system path.
  • [CREDENTIALS_UNSAFE]: The skill accesses the GPC_SERVICE_ACCOUNT environment variable to authenticate with Google Play services, representing a standard exposure of sensitive configuration for API access.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 12, 2026, 01:58 PM