zadig

WarnAudited by ClawScan on May 10, 2026.

Overview

This is mostly a coherent Zadig API client, but it contains under-disclosed shell execution and can send a powerful Zadig token over non-HTTPS connections.

Review index.js before installing. Only use an HTTPS Zadig URL, create a least-privilege API token, and require manual confirmation for project deletion, deployments, cluster changes, and user or role changes. The shell execution path should be removed or fully explained and constrained before trusting this skill.

Findings (3)

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.

High

#ASI05: Unexpected Code Execution

What this means

If that code path is invoked, the skill can run local shell commands beyond the stated API-client behavior.

Why it was flagged

The code contains synchronous shell command execution, while the skill documentation and package permissions only disclose environment-variable access and network calls to Zadig.

Skill content

output = execSync(cmd, { encoding: 'utf8', timeout: 15000 });

Recommendation

Remove the shell execution or replace it with direct API calls. If it is truly required, document it clearly, constrain allowed commands, sanitize all arguments, and require explicit user approval.

High

#ASI03: Identity and Privilege Abuse

What this means

A misconfigured http:// Zadig URL could expose the API token and any DevOps permissions it grants.

Why it was flagged

The client chooses the plain HTTP library for non-HTTPS URLs while still sending the Zadig bearer token, despite the package declaring network:https and describing HTTPS-only calls.

Skill content

const isHttps = url.protocol === 'https:';
const lib = isHttps ? https : http;
...
'Authorization': 'Bearer ' + API_KEY

Recommendation

Enforce HTTPS by rejecting non-https ZADIG_API_URL values, and use a least-privilege Zadig token.

Medium

#ASI02: Tool Misuse and Exploitation

What this means

Mistaken or over-broad agent use could delete projects, trigger or cancel workflows, change clusters, or modify users and roles.

Why it was flagged

The skill exposes destructive and administrative DevOps operations. This matches the stated DevOps automation purpose, but these actions can materially change deployments and account permissions.

Skill content

| 项目 | 创建/查询/删除项目 |
| 工作流 | 触发/查询/取消/重试/审批工作流 |
| 集群 | 管理 Kubernetes 集群 |
| 权限 | 用户、角色、成员管理 |

Recommendation

Use a scoped token, set a default project/environment where possible, and require explicit confirmation before destructive, deployment, or permission-management actions.