dmux-workflows

Pass

Audited by Gen Agent Trust Hub on Apr 14, 2026

Risk Level: SAFEPROMPT_INJECTIONCOMMAND_EXECUTIONEXTERNAL_DOWNLOADS
Full Analysis
  • [PROMPT_INJECTION]: The orchestration patterns create an indirect prompt injection surface.
  • Ingestion points: The skill merges output from parallel agent panes back into a shared context and ingests task definitions from a plan.json file.
  • Boundary markers: No markers are defined to isolate untrusted agent output during the merge process.
  • Capability inventory: The skill possesses file system access via git and shell execution capabilities via the orchestrate-worktrees.js helper.
  • Sanitization: There is no evidence of sanitization for data transferred between agent sessions.
  • [COMMAND_EXECUTION]: The orchestration helper script scripts/orchestrate-worktrees.js performs dynamic shell command execution based on the launcherCommand field in plan.json. This could be exploited if an agent is tricked into generating or modifying the configuration file.
  • [EXTERNAL_DOWNLOADS]: The skill directs users to download and install the dmux tool from an external GitHub repository (github.com/standardagents/dmux) which is not among the verified technology providers.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 14, 2026, 12:19 AM
Security Audit — agent-trust-hub — dmux-workflows