Feishu Media

• Outcome: Pass
• Latest audit: May 29, 2026 at 8:29 PM UTC
• Version:
  1.1.1

Overview

This is a Feishu media-sending helper whose file uploads, conversions, and API calls are disclosed and aligned with its purpose, but users must handle recipients, files, and Feishu credentials carefully.

Install only if you want the agent to send selected files or media to Feishu. Confirm the exact recipient and file before each send, avoid packaging whole folders unless you have reviewed the contents, do not paste Feishu app secrets into casual prompts or logs, and use a least-privilege Feishu app for uploads.

SkillSpector

By NVIDIA

Vulnerability Patterns

• Data Exfiltration: External Transmission, Env Variable Harvesting, File System Enumeration
• Excessive Agency: Unrestricted Tool Access, Autonomous Decision Making, Scope Creep
• Trigger Abuse: Overly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
• Prompt Injection: Instruction Override, Hidden Instructions, Exfiltration Commands
• Privilege Escalation: Excessive Permissions, Sudo/Root Execution, Credential Access

Findings (6)

Context-Inappropriate Capability

Medium

• Confidence: 94% confidence
• Finding: The skill goes beyond using the declared `message` tool and instructs operators to run arbitrary shell commands and direct `curl` calls with `app_id`/`app_secret` to obtain tokens and upload files. This expands the skill from simple media sending into credential handling and external API execution, increasing the risk of secret exposure, unreviewed outbound transfers, and command misuse.

Context-Inappropriate Capability

Medium

• Confidence: 82% confidence
• Finding: Documenting `ffmpeg`/`ffprobe` conversion and probing introduces local command execution and file processing capabilities not inherent to a send-only skill. Even if intended for compatibility, this broadens the operational scope and can lead to unsafe handling of local files or dependency abuse in environments where shell execution should not be encouraged.

Context-Inappropriate Capability

Low

• Confidence: 78% confidence
• Finding: The archive creation guidance adds file-packaging behavior beyond straightforward media sending and can facilitate bundling multiple local files for outbound transmission. In a messaging context, that increases the chance of over-collection and accidental exfiltration of unrelated files.

Vague Triggers

Medium

• Confidence: 76% confidence
• Finding: The activation description is broad enough to match many ordinary requests about sending files or media, which raises the risk of accidental invocation. In a skill that can transmit local or remote content externally, mis-triggering materially increases exposure risk.

Missing User Warnings

Medium

• Confidence: 91% confidence
• Finding: The documentation explains how to send local files and remote media to Feishu but does not warn that this is an external transmission of potentially sensitive data. Without clear disclosure and confirmation requirements, users may inadvertently exfiltrate confidential files, images, or URLs.

Missing User Warnings

High

• Confidence: 97% confidence
• Finding: The video workflow instructs use of `appId` and `appSecret` to mint access tokens and upload files, but it omits strong warnings about credential sensitivity and outbound data transfer. This combination creates a significant risk of secret leakage, unauthorized API use, and transmission of sensitive media to unintended recipients.

VirusTotal

65/65 vendors flagged this skill as clean.