follow-up-after-call
Pass
Audited by Gen Agent Trust Hub on Jun 16, 2026
Risk Level: SAFE
Full Analysis
- Human-in-the-Loop Controls: The skill is designed with mandatory checkpoints requiring the user to review and approve content before it is converted into external drafts or messages. This ensures that the user maintains final authority over all communications and prevents automated actions.
- Restricted Data Operations: Instructions explicitly limit the skill's capabilities to prevent unauthorized modifications to external systems. It is directed to use 'copy/paste' only for CRM updates and is prohibited from sending emails or posting to internal channels without direct user intervention.
- Input Validation and Grounding: The procedure requires 'grounded call evidence' and uses a structured resolution process to clarify ambiguous inputs. This approach ensures that generated summaries and follow-up packages are based on verified source data rather than inferred or generated details.
- Data Processing Considerations: The skill processes information from call transcripts and message threads, which are external data sources. It manages the associated risks by using structured formatting, such as Markdown block quotes for drafts, and clear placeholders for missing or ambiguous information.
Audit Metadata