clari-core-workflow-a
Pass
Audited by Gen Agent Trust Hub on Sep 13, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process data from the external Clari API via the
export_and_downloadmethod. This creates a surface for indirect prompt injection, where malicious instructions could theoretically be embedded in source data fields (e.g., owner names or CRM notes) to influence the agent's behavior during transformation or logging. - Ingestion points: Data is fetched from the Clari API in
export_pipeline.pywithin therun_exportfunction. - Boundary markers: None detected in the provided code snippets.
- Capability inventory: The skill possesses the capability to write to a Snowflake database and print data to standard output.
- Sanitization: The skill uses parameterized queries (
%(field)s) for SQLINSERTstatements, which effectively prevents SQL injection for the data payload, though the table creation uses string interpolation for the table name.
Audit Metadata