skills/testland/qa/atlas-migrations/Gen Agent Trust Hub

atlas-migrations

Fail

Audited by Gen Agent Trust Hub on Aug 12, 2026

Risk Level: HIGHEXTERNAL_DOWNLOADSREMOTE_CODE_EXECUTIONCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill fetches an installation script from the official Atlas project domain at https://atlasgo.sh.- [REMOTE_CODE_EXECUTION]: The downloaded script is executed via 'sh install-atlas.sh' to install the Atlas CLI tool. This is a standard and documented setup procedure for this well-known service.- [COMMAND_EXECUTION]: The skill utilizes the 'atlas' command-line tool to perform various database operations, including applying schemas, generating migration diffs, and linting migration files.- [PROMPT_INJECTION]: The skill processes external database schema definitions in SQL or HCL format, which creates a surface for indirect prompt injection. If these files contain hidden instructions, they could influence the agent's behavior.
  • Ingestion points: The skill reads local files such as 'schema.sql' and files within the 'migrations/' directory.
  • Boundary markers: The instructions do not implement specific delimiters or 'ignore' directives to prevent the agent from obeying instructions embedded in the schema data.
  • Capability inventory: The skill enables the agent to read and write files and perform database modifications using the CLI.
  • Sanitization: No content validation or filtering is specified for the schema files before they are processed by the agent.
Recommendations
  • HIGH: Downloads and executes remote code from: https://atlasgo.sh - DO NOT USE without thorough review
Audit Metadata
Risk Level
HIGH
Analyzed
Aug 12, 2026, 10:52 AM
Security Audit — agent-trust-hub — atlas-migrations