restate-workflows

Pass

Audited by Gen Agent Trust Hub on Mar 18, 2026

Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill provides guidance and a TypeScript code sample in SKILL.md for invoking the joelclaw CLI using the node:child_process.spawn method. This functionality is part of the intended purpose to bridge external repositories with the joelclaw runtime.\n- [PROMPT_INJECTION]: The skill implements a communication pattern that includes a prompt field for task descriptions, creating an indirect prompt injection surface.\n
  • Ingestion points: The task.prompt and task.kind fields in the JSON payload structure defined in SKILL.md.\n
  • Boundary markers: Absent; the instructions do not specify the use of delimiters or safety instructions to prevent the agent from obeying instructions embedded within the data.\n
  • Capability inventory: Execution of system commands via joelclaw queue emit and joelclaw workload run as described in SKILL.md.\n
  • Sanitization: Absent for prompt content; the provided code ensures safe argument passing via JSON stringification but does not filter or sanitize the contents of the prompt itself.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 18, 2026, 12:38 PM