Back to skill

Security audit

Huozi

Security checks for vulnerabilities and agentic risk

Overview

This skill matches its publishing purpose, but its Bash API examples create a real local command-injection risk and its delete operation is under-scoped.

Install only if you trust Huozi with the email and content you publish, and use extra care around generated Bash commands: API requests should serialize JSON safely instead of pasting user content into shell-quoted curl strings. Confirm the exact page slug before any delete request.

Vulnerability Patterns
  • Insecure Skill Coding PracticesFinds exploitable flaws such as hardcoded secrets or command injection
  • Skill Instruction HijackingAlters the agent's session goals or safety constraints when the skill loads
  • Agent Memory PoisoningWrites attacker-controlled rules into memory that affect later sessions
  • Remote Payload Retrieval and ExecutionFetches external code whose behavior can change after review
  • Embedded Malicious CodeShips malicious scripts inside the skill and executes them locally
Findings (1)

T09 · Insecure Skill Coding Practices

Error
Location
SKILL.md:22
Finding

Shell Command Injection Through Unsafe Interpolation into curl Commands

Content
View full analysis
"}' ``` Verification request at `SKILL.md:34-38`: ```bash curl -s -X POST https://huozi.app/api/v1/auth/verify \ -H "Content-Type: application/json" \ -d '{"email": "", "code": ""}' ``` Workspace setup request at `SKILL.md:48-52`: ```bash curl -s -X POST https://huozi.app/api/v1/auth/setup \ -H "Authorization: Bearer " \ -H "Content-Type: application/json" \ -d '{"workspace_slug": ""}' ``` Markdown publishing request at `SKILL.md:67-70`: ```bash curl -s -X POST https://huozi.app/api/v1/pages \ -H "Authorization: Bearer " \ -H "Content-Type: application/json" \ -d '{"title": "", "slug": "<slug>", "content": "<markdown>"}' ``` HTML publishing request at `SKILL.md:78-81`: ```bash curl -s -X POST https://huozi.app/api/v1/pages \ -H "Authorization: Bearer <api_key>" \ -H "Content-Type: application/json" \ -d '{"title": "<title>", "slug": "<slug>", "content": "<html>", "content_type": "html"}' ``` ### Technical Analysis The skill instructs the agent to substitute user-controlled email addresses, verification codes, workspace slugs, titles, page slugs, Markdown, and HTML directly into JSON enclosed by shell single quotes. In POSIX-compatible shells, a single quote contained in an interpolated value terminates the quoted argument. Subsequent characters can consequently be interpreted as shell operators or commands rather than request data. This construction is also not a valid general-purpose JSON serialization mechanism ...[truncated 2339 chars]
Remediation
View remediation
Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Findings (3)

Credential Access

High
Category
Privilege Escalation
Confidence
70% confidence
Finding

Code accesses credential files (SSH keys, AWS credentials, etc.). This could indicate credential theft attempts.

Content

Scanner excerpt · SKILL.md (reported line 91)May include surrounding context.

text

- Set `content_type` to `"html"` (defaults to `"markdown"` if omitted)
- Same slug = upsert, versioning, and access tokens work identically to Markdown pages

### HTML Input Format

External Transmission

Medium
Category
Data Exfiltration
Confidence
93% confidence
Finding

The skill instructs the agent to immediately send the user's email address to an external third-party service as part of an onboarding flow. While this is expected for account registration, it is still an external transmission of personal data and can create privacy and consent risks if the user is not clearly informed before the transfer.

Content

Scanner excerpt · SKILL.md (reported line 22)May include surrounding context.

Tell the user: "Let's set up your Huozi account. What's your email?" Then call:

bash
curl -s -X POST https://huozi.app/api/v1/auth/signup \
  -H "Content-Type: application/json" \
  -d '{"email": "<user_email>"}'

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
88% confidence
Finding

The skill exposes a destructive delete capability but provides no requirement for explicit user confirmation, no warning about irreversibility, and no safety checks before deletion. In an agentic setting, this increases the chance of accidental or ambiguous deletion requests resulting in permanent content loss on a remote service.

Content

No source excerpt is available for this finding.