etl-pipelines
Pass
Audited by Gen Agent Trust Hub on Mar 5, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill's instructions and examples do not contain any malicious patterns such as direct prompt injection, unauthorized data access, or obfuscation.
- [CREDENTIALS_UNSAFE]: The skill follows security best practices by advising the use of environment variables for authentication tokens ($OPENCLAW_API_KEY) and cloud credentials ($AWS_ACCESS_KEY_ID) rather than hardcoding them.
- [COMMAND_EXECUTION]: All command-line examples use the vendor's own 'openclaw' CLI tool for legitimate pipeline management tasks like creating and running ETL jobs.
- [DATA_EXFILTRATION]: No evidence of data being sent to unauthorized external domains was found; network operations are limited to configured sources and targets such as PostgreSQL, S3, and BigQuery.
- [PROMPT_INJECTION]:
- Ingestion points: Data enters the agent context via configuration files (e.g., sales_config.json) and external data sources (CSV, JSON, SQL databases).
- Boundary markers: Absent; the skill does not explicitly define delimiters for external data content.
- Capability inventory: Includes file system access (S3/local), database connections, and execution of SQL queries or Pandas operations.
- Sanitization: The skill identifies 'schema validation' as a key capability, which helps ensure data integrity.
Audit Metadata