draft-outreach
Pass
Audited by Gen Agent Trust Hub on Nov 11, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted data from multiple external sources and uses that content to influence the generation of outreach emails.
- Ingestion points: Data is ingested from Salesforce GraphQL API responses (Contacts, Accounts, Opportunities, Tasks, Events), and from external Slack and document search tools (Step 2 and 2b in
SKILL.md). - Boundary markers: The skill lacks explicit delimiters or specific 'ignore embedded instructions' directives when processing the body of retrieved Slack messages or document notes during the drafting phase.
- Capability inventory: The agent has the capability to create email drafts and perform data writes to Salesforce via the
dispatchtool. - Sanitization: There are no explicit instructions for the agent to sanitize or validate the content of retrieved records before interpolating them into the final draft.
Audit Metadata