dmux-workflows
Pass
Audited by Gen Agent Trust Hub on Jun 20, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill instructs the user to install the
dmuxpackage globally from the NPM registry usingnpm install -g dmuxand references an external repository atgithub.com/standardagents/dmuxfor the tool's source code. - [COMMAND_EXECUTION]: The skill guides the user to execute shell commands for session management and parallelization, including
dmux,tmux capture-pane, andgit worktreeoperations to isolate agent environments. - [PROMPT_INJECTION]: The skill exhibits a surface for indirect prompt injection due to its orchestration of multi-agent workflows:
- Ingestion points: Parallel agent panes ingest prompts and data (e.g., Pane 1's research findings are fed into Pane 2's implementation context as described in 'Pattern 1').
- Boundary markers: The instructions do not define clear delimiters or include 'ignore embedded instructions' warnings when merging output from one agent session into another.
- Capability inventory: The skill facilitates broad capabilities including shell execution via tmux panes and file system modifications via git and agent harnesses across the user's workspace.
- Sanitization: There is no explicit sanitization, validation, or filtering of the content produced by sub-agents before it is processed by the main session or other parallel panes.
Audit Metadata