Claude Code tmux Runner
PassAudited by VirusTotal on May 12, 2026.
Overview
Type: OpenClaw Skill Name: claude-tmux-runner Version: 1.0.0 The skill is classified as suspicious due to the potential for shell injection. The `SKILL.md` instructs the agent to pass user-provided strings, such as 'Your task description' and '[task-name]', directly as arguments to the `~/.openclaw/scripts/claude-tmux.sh` script. While the script itself is not provided for analysis, this pattern creates a significant vulnerability for shell injection if the script does not properly sanitize or escape these user inputs before execution, potentially allowing an attacker to execute arbitrary commands. There is no evidence of intentional malicious behavior like data exfiltration or backdoor installation.
Findings (0)
Artifact-based informational review of SKILL.md, metadata, install specs, static scan signals, and capability signals. ClawScan does not execute the skill or run runtime probes.
A task prompt may be passed into a shell command that starts a background process, so unexpected helper-script behavior could affect the local environment.
The skill is designed to run a local shell helper with user-provided task text. This is central to the purpose, but users should be aware it executes local commands.
tools: Bash, Read, Write ... ~/.openclaw/scripts/claude-tmux.sh start "Your task description" [task-name]
Approve each task before launch, keep task arguments properly quoted, and inspect the helper script before relying on it.
The safety of the main executable depends on a local script that was not included for review.
The supplied package does not include installation steps or the referenced ~/.openclaw/scripts/claude-tmux.sh implementation, so the actual script provenance is outside the reviewed artifacts.
No install spec — this is an instruction-only skill.
Verify the contents, source, and permissions of ~/.openclaw/scripts/claude-tmux.sh before running any task.
Claude Code tasks may continue running after the main chat returns, consuming resources or modifying generated files until stopped.
The skill intentionally starts long-running background Claude Code tasks. This is disclosed and includes stop commands, so it is a notice rather than a concern.
Run multiple Claude Code tasks in parallel using tmux sessions. Non-blocking, background execution with status monitoring.
Use the documented status and stop/stop-all commands to monitor and end background jobs when they are no longer needed.