planetscale-pscale-cli-automation

Pass

Audited by Gen Agent Trust Hub on Sep 3, 2026

Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill provides instructions for the agent to execute the pscale command-line tool for authentication (pscale auth check), resource management (pscale database list), and SQL query execution (pscale sql).
  • [EXTERNAL_DOWNLOADS]: Fetches additional setup scripts and skill definitions from PlanetScale's official GitHub repository (github.com/planetscale/skills). These are vendor-owned resources used for expanding the agent's capabilities within the PlanetScale ecosystem.
  • [INDIRECT_PROMPT_INJECTION]: The skill processes structured JSON output from the CLI (pscale agent-guide --format json).
  • Ingestion points: CLI output from pscale subcommands (SKILL.md).
  • Boundary markers: Explicitly mandates the use of --format json to ensure structured data parsing rather than natural language interpretation.
  • Capability inventory: The skill allows for database management and SQL execution via the pscale tool (SKILL.md).
  • Sanitization: Relies on structured JSON parsing to separate command results from executable logic.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 3, 2026, 02:08 PM
Security Audit — agent-trust-hub — planetscale-pscale-cli-automation