Back to skill

Security audit

Zernio API

Security checks for vulnerabilities and agentic risk

Overview

ClawScan rates this skill suspicious with medium confidence: it documents Zernio API operations, but reports some capabilities outside the stated reference purpose and an install command using a mutable package version.

The scanner recommends against installation, and the verdict places this in the suspicious Review bucket; it does not label the author malicious. Before installing, consider the mutable installer version, the reported scope mismatch, and the unreviewed behavior in the two partially analyzed files. If using the API examples, review the destination and data involved and take care with actions that publish, message, upload, or share credentials.

Vulnerability Patterns
  • 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
  • Unauthorized Access and Privilege EscalationObtains permissions beyond the task's legitimate needs
Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Output HandlingUnvalidated Output Injection, Cross-Context Output, Unbounded Output
  • Tool MisuseTool Parameter Abuse, Chaining Abuse, Unsafe Defaults
  • MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
Findings (51)

Description-Behavior Mismatch

High
Category
MCP Tool Poisoning
Confidence
97% confidence
Finding

The supplied chunk performs skill installation, which is materially different from the declared documentation purpose. Running it can fetch and execute external package code. The contents and behavior of the installed skill are not supplied and cannot be verified.

Content

No source excerpt is available for this finding.

Description-Behavior Mismatch

High
Category
MCP Tool Poisoning
Confidence
96% confidence
Finding

The commands target contact data rather than scheduled posts, advertising, or supporting API infrastructure. Although they use the declared Zernio service, contact-field management is a distinct capability not disclosed in the description.

Content

No source excerpt is available for this finding.

Description-Behavior Mismatch

High
Category
MCP Tool Poisoning
Confidence
92% confidence
Finding

The supplied request creates a contact rather than scheduling a post or managing an advertisement. Contact management is a distinct capability not disclosed in the description, and the snippet does not establish it as a supporting step for either declared function. The request targets the declared service, but that alone does not make the additional capability covered.

Content

No source excerpt is available for this finding.

Description-Behavior Mismatch

High
Category
MCP Tool Poisoning
Confidence
91% confidence
Finding

The command uses an inbox messaging endpoint rather than a post-scheduling or advertising endpoint. Initiating a conversation with a specific recipient is a distinct capability not clearly disclosed by the declared purpose, rather than a supporting implementation detail. The command contains placeholder credentials, so the snippet alone does not establish that a message was actually sent.

Content

No source excerpt is available for this finding.

Incomplete referenced artifact analysis

High
Category
analysis-evasion
Confidence
100% confidence
Finding

Referenced artifact was not completely inspected

Content

Scanner excerpt · SKILL.md (reported line 39)May include surrounding context.

md
- [rules/posts.md](rules/posts.md) - Create, schedule, retry, edit, unpublish, bulk upload

Incomplete referenced artifact analysis

High
Category
analysis-evasion
Confidence
100% confidence
Finding

Referenced artifact was not completely inspected

Content

Scanner excerpt · SKILL.md (reported line 42)May include surrounding context.

md
- [rules/queue.md](rules/queue.md) - Queue management, slots configuration

Tool Parameter Abuse

High
Category
Tool Misuse
Confidence
80% confidence
Finding

Tool parameters are crafted to achieve unintended or unsafe behavior. Parameter abuse can bypass intended safety checks (e.g. shell=True, --force, dangerous glob patterns).

Content

Scanner excerpt · rules/gmb.md (reported line 37)May include surrounding context.

-d '{"locationNames": ["accounts/123/locations/456", "accounts/123/locations/789"], "pageSize": 50}'

text

Reply to a review via the unified inbox: `POST /v1/inbox/reviews/{reviewId}/reply` (see `inbox.md`). Delete a reply (GBP only): `DELETE /v1/inbox/reviews/{reviewId}/reply`.

## Location Details

MCP server referenced without pinned version: 'npx clawhub'.

Medium
Category
MCP Rug Pull
Confidence
88% confidence
Finding

The installation command uses npx with a mutable @latest version, allowing newly published package code to execute on the installer's machine without a reviewed version boundary. This is a supply-chain exposure rather than evidence of malicious behavior or an MCP server; exploitation would require compromise of the package or its publication channel.

Content

No source excerpt is available for this finding.

External Transmission

Medium
Category
Data Exfiltration
Confidence
60% confidence
Finding

Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.

Content

Scanner excerpt · README.md (reported line 54)May include surrounding context.

Quick Example

typescript
const response = await fetch('https://zernio.com/api/v1/posts', {
  method: 'POST',
  headers: {
    'Authorization': `Bearer ${process.env.ZERNIO_API_KEY}`,

External Transmission

Medium
Category
Data Exfiltration
Confidence
60% confidence
Finding

Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.

Content

Scanner excerpt · SKILL.md (reported line 18)May include surrounding context.

bash
# 1. Create profile
curl -X POST https://zernio.com/api/v1/profiles \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -d '{"name": "My Brand"}'

The Quick Start publishes a post immediately without warning that it will go live on the connected account.

Medium
Category
Output Handling
Confidence
94% confidence
Finding

The skill's own Quick Start instructs the user to create a post with publishNow: true, which immediately publishes content rather than merely creating a draft or scheduling it. The step is labeled only "Create post" and provides no warning about the externally visible effect.

Content

No source excerpt is available for this finding.

External Transmission

Medium
Category
Data Exfiltration
Confidence
60% confidence
Finding

Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.

Content

Scanner excerpt · rules/account-groups.md (reported line 15)May include surrounding context.

Group names must be unique per user — POST and PUT return 409 on collision.

bash
curl -X POST https://zernio.com/api/v1/account-groups \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -d '{
    "name": "Marketing Accounts",

External Transmission

Medium
Category
Data Exfiltration
Confidence
60% confidence
Finding

Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.

Content

Scanner excerpt · rules/accounts.md (reported line 48)May include surrounding context.

TikTok requires fetching creator-specific settings before posting — use this to populate platformSpecificData correctly (allowed privacy levels, whether the creator disabled duet/stitch, max video duration).

bash
curl "https://zernio.com/api/v1/accounts/ACCOUNT_ID/tiktok/creator-info" \
  -H "Authorization: Bearer YOUR_API_KEY"

External Transmission

Medium
Category
Data Exfiltration
Confidence
60% confidence
Finding

Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.

Content

Scanner excerpt · rules/ads.md (reported line 37)May include surrounding context.

bash
# Defaults to last 90 days; max 90-day range
curl "https://zernio.com/api/v1/ads?platform=facebook&status=active&limit=50" \
  -H "Authorization: Bearer YOUR_API_KEY"

External Transmission

Medium
Category
Data Exfiltration
Confidence
60% confidence
Finding

Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.

Content

Scanner excerpt · rules/ads.md (reported line 88)May include surrounding context.

Creates the full campaign > ad set > ad hierarchy with custom creative.

bash
curl -X POST https://zernio.com/api/v1/ads/create \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{

Customer-list upload instructions lack a privacy warning about transmitting customer contact details to external services.

Medium
Category
Output Handling
Confidence
86% confidence
Finding

The upload example sends customer email addresses and phone numbers to Zernio for advertising audience matching without warning the user about the privacy implications or requiring authorization to share those records. The server-side hashing note does not explain that Zernio receives the plaintext identifiers before hashing.

Content

No source excerpt is available for this finding.

External Transmission

Medium
Category
Data Exfiltration
Confidence
60% confidence
Finding

Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.

Content

Scanner excerpt · rules/api-keys.md (reported line 14)May include surrounding context.

Create

bash
curl -X POST https://zernio.com/api/v1/api-keys \
  -H "Authorization: Bearer YOUR_MASTER_API_KEY" \
  -d '{
    "name": "Analytics Read-Only",

External Transmission

Medium
Category
Data Exfiltration
Confidence
60% confidence
Finding

Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.

Content

Scanner excerpt · rules/broadcasts.md (reported line 28)May include surrounding context.

bash
# 1. Create draft (Telegram example — text + segmentFilters for auto-recipient)
curl -X POST https://zernio.com/api/v1/broadcasts \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -d '{
    "profileId": "PROFILE_ID",

The broadcast workflow sends or schedules bulk messages without warning about external delivery and its irreversible effects.

Medium
Category
Output Handling
Confidence
93% confidence
Finding

The typical flow instructs the agent to invoke the send or schedule endpoint, rather than merely listing available APIs. It provides no warning that the action delivers content to the selected recipients through external messaging platforms, or that cancelling a broadcast does not retract messages already delivered.

Content

No source excerpt is available for this finding.

External Transmission

Medium
Category
Data Exfiltration
Confidence
60% confidence
Finding

Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.

Content

Scanner excerpt · rules/connect.md (reported line 40)May include surrounding context.

bash
# Get OAuth URL
curl "https://zernio.com/api/v1/connect/twitter?profileId=PROFILE_ID&callbackUrl=https%3A%2F%2Fyourapp.com%2Fcallback" \
  -H "Authorization: Bearer YOUR_API_KEY"

The Bluesky connection flow sends an app password to Zernio without warning the user about third-party credential disclosure.

Medium
Category
Output Handling
Confidence
92% confidence
Finding

The connection example posts the user's Bluesky identifier and app password to zernio.com rather than directly to Bluesky. Although the request body shows the fields, the workflow does not warn users that a third-party service will receive their credential.

Content

No source excerpt is available for this finding.

External Transmission

Medium
Category
Data Exfiltration
Confidence
60% confidence
Finding

Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.

Content

Scanner excerpt · rules/contacts.md (reported line 22)May include surrounding context.

List

bash
curl "https://zernio.com/api/v1/contacts?profileId=PROFILE_ID&tag=vip&platform=whatsapp&search=anna&limit=100" \
  -H "Authorization: Bearer YOUR_API_KEY"

External Transmission

Medium
Category
Data Exfiltration
Confidence
60% confidence
Finding

Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.

Content

Scanner excerpt · rules/contacts.md (reported line 82)May include surrounding context.

bash
# Plain text field
curl -X POST https://zernio.com/api/v1/custom-fields \
  -d '{"profileId": "...", "name": "Job Title", "type": "text"}'

# Select (dropdown) — requires options[]

External Transmission

Medium
Category
Data Exfiltration
Confidence
60% confidence
Finding

Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.

Content

Scanner excerpt · rules/gmb.md (reported line 24)May include surrounding context.

| POST | /v1/accounts/{accountId}/gmb-reviews/batch | Fetch reviews across multiple locations in one request |

bash
curl "https://zernio.com/api/v1/accounts/ACC/gmb-reviews?pageSize=50" \
  -H "Authorization: Bearer YOUR_API_KEY"

External Transmission

Medium
Category
Data Exfiltration
Confidence
60% confidence
Finding

Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.

Content

Scanner excerpt · rules/gmb.md (reported line 52)May include surrounding context.

bash
# Update business hours + holiday closures
curl -X PUT https://zernio.com/api/v1/accounts/ACC/gmb-location-details \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -d '{
    "updateMask": "regularHours,specialHours",