Security audit

Fast Image

Security checks across malware telemetry and agentic risk

Overview

The skill does what it says, but its image-sending helper can run unintended local shell commands if an image path, channel, target, or message contains shell metacharacters.

Install only if you understand it sends local image files to external channels. Avoid using untrusted filenames, channel names, targets, or message text until the shell invocation is fixed; the safer implementation would call openclaw without shell:true and validate destinations.

SkillSpector

By NVIDIA

Vulnerability Patterns

Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access

Findings (4)

Context-Inappropriate Capability

Medium

Confidence: 95% confidence
Finding: The script invokes `openclaw` via `spawn(..., { shell: true })` while passing user-controlled values such as `channel`, `target`, `message`, and an image-derived path. Using a shell here creates command-injection risk if any argument contains shell metacharacters, and it also expands the skill from simple file handling into external message transmission.

Vague Triggers

Medium

Confidence: 89% confidence
Finding: The keyword list is broad enough that the skill may be invoked for generic image- or media-related requests beyond the user's actual intent. Because this skill copies local files and sends them to external channels, unintended invocation increases the chance of accidental disclosure of local images or metadata.

Missing User Warnings

Medium

Confidence: 91% confidence
Finding: The skill documentation does not clearly warn users that it copies files into a media directory, may create temporary compressed versions, and deletes temp files after sending. Missing disclosure of these filesystem side effects can lead to users unintentionally exposing sensitive local images, retaining copies in unexpected locations, or losing temporary artifacts needed for recovery or audit.

Missing User Warnings

Medium

Confidence: 86% confidence
Finding: The skill sends a local image to an external channel with no in-script disclosure, confirmation, or safeguard beyond command-line arguments. In an agent skill context, silent exfiltration of local files is security-relevant because users may not realize a local path is being transmitted off-host.

VirusTotal

64/64 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

Detected: suspicious.dangerous_exec

Shell command execution detected (child_process).

Critical

Code: suspicious.dangerous_exec
Location: send_image.mjs:75