ClawHub
Back to skill

Security audit

huawei-cloud-swr-image-management

Security checks across malware telemetry and agentic risk

Overview

The skill is coherent for Huawei Cloud SWR administration, but it needs Review because it handles durable registry credentials and destructive cloud operations while documenting unsafe secret handling.

Install only if you are comfortable giving the agent Huawei Cloud SWR management authority. Use least-privilege IAM, confirm every create/update/delete command, prefer temporary login tokens, avoid printing or passing registry passwords on command lines, and treat one-year CreateSecret credentials as production secrets that need secure storage and rotation.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Findings (3)

Description-Behavior Mismatch

Medium
Confidence
93% confidence
Finding
The skill manifest frames credential support as obtaining Docker login credentials, but this troubleshooting guide additionally instructs users to create long-term SWR credentials via `CreateSecret` valid for one year. That expands the operational scope from short-lived auth assistance into durable credential issuance, increasing the chance of privilege persistence, secret sprawl, and misuse in CI/CD environments if users follow the guidance without stronger safeguards.

Intent-Code Divergence

Medium
Confidence
88% confidence
Finding
The guide correctly warns against echoing cloud credentials, but later instructs users to decode a Docker auth token and pass the resulting username and password directly on the command line. Command-line secrets can be exposed through shell history, process listings, terminal logs, CI job logs, and agent transcript capture, which creates a real secret-handling vulnerability even if the token is temporary.

Intent-Code Divergence

Medium
Confidence
95% confidence
Finding
The documentation inaccurately states that deleting a tag permanently removes the image version, while also acknowledging that other tags referencing the same digest are unaffected. This semantic contradiction can mislead operators into making unsafe lifecycle decisions, causing accidental retention of deployable images or mistaken assumptions about actual deletion behavior during cleanup, rollback, or incident response.

VirusTotal

VirusTotal findings are pending for this skill version.

View on VirusTotal

Static analysis

Detected: suspicious.exposed_secret_literal

File appears to expose a hardcoded API secret or token.

Critical
Code
suspicious.exposed_secret_literal
Location
references/task-auth-management.md:83