workflow-rig

Pass

Audited by Gen Agent Trust Hub on Mar 18, 2026

Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill facilitates the execution of arbitrary shell commands and file system operations within a Kubernetes environment through the joelclaw CLI handlers.
  • Evidence: The 'shell' and 'infer' handlers are documented to support bash execution, file editing, and git operations (clone, commit, push) as part of the workload execution process.
  • [PROMPT_INJECTION]: The skill presents an indirect prompt injection surface due to its architecture of ingesting external data to drive agent behavior.
  • Ingestion points: Untrusted data enters the agent context through user-provided intent strings in planning commands and external stage definitions provided via the '--stages-from' argument.
  • Boundary markers: The skill documentation does not mention the use of specific delimiters or markers to separate instructions from data.
  • Capability inventory: The system has broad capabilities, including git repository access, file system modification, and shell command execution.
  • Sanitization: There is no description of sanitization or validation of input content before it is processed by the agentic 'infer' or 'shell' handlers.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 18, 2026, 12:38 PM
Security Audit — agent-trust-hub — workflow-rig