skills/himanich/skills/aem-workflow/Gen Agent Trust Hub

aem-workflow

Pass

Audited by Gen Agent Trust Hub on Apr 20, 2026

Risk Level: SAFECREDENTIALS_UNSAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [DATA_EXPOSURE_AND_EXFILTRATION]: Hardcoded default credentials (admin:admin) are utilized in command-line examples illustrating the use of the AEM Workflow REST API.
  • Evidence: Examples found in workflow-triggering/SKILL.md, workflow-triggering/references/workflow-triggering/triggering-mechanisms.md, and workflow-orchestrator/references/workflow-foundation/quick-start-guide.md.
  • [DYNAMIC_EXECUTION]: The Granite Workflow Engine evaluates ECMA (JavaScript) script strings at runtime for transition rules and loop-back logic.
  • Evidence: Patterns documented in workflow-model-design/references/workflow-model-design/step-types-catalog.md and workflow-model-design/references/workflow-model-design/model-design-patterns.md.
  • [INDIRECT_PROMPT_INJECTION]: The skill implements Java components and scripts that ingest and process potentially untrusted data from the JCR repository (such as payloads and metadata) without explicit sanitization guidance.
  • Ingestion points: Payload paths and MetaDataMap properties accessed within WorkflowProcess and ParticipantStepChooser implementations (e.g., in workflow-development/SKILL.md).
  • Boundary markers: Absent in provided Java and XML templates.
  • Capability inventory: The skill provides capabilities for JCR write operations, workflow lifecycle management (start/stop/complete), and script execution.
  • Sanitization: No input validation or content sanitization logic is demonstrated in the provided code templates.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 20, 2026, 10:52 AM