Back to skill

Security audit

AI4Trade Full OpenClaw

Security checks for vulnerabilities and agentic risk

Overview

The skill is clear about AI4Trade trading workflows, but it recommends unverified remote skill and plugin installation paths that could change what gets installed after review.

Install only if you trust AI4Trade and the @clawtrader plugin publisher. Prefer the reviewed local bundle, do not auto-fetch replacement skill files from remote URLs, pin and review any plugin before enabling it, store AI4Trade tokens in a secret manager, require exact approval before publishing/following/copy-trading, and use only wss:// for WebSocket monitoring.

Vulnerability Patterns
  • Remote Payload Retrieval and ExecutionFetches external code whose behavior can change after review
  • Insecure DependenciesIntroduces malicious components through unsafe dependency sources
  • 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
Findings (3)

T03 · Remote Payload Retrieval and Execution

Error
Location
copytrade/SKILL.md:23
Finding

Mutable Remote Skill Content Can Be Retrieved and Installed Without Integrity Verification

Content
View full analysis
Remediation
View remediation

T08 · Insecure Dependencies

Error
Location
copytrade/SKILL.md:45
Finding

Third-Party OpenClaw Plugins Are Installed Without Version or Integrity Pinning

Content
View full analysis
Remediation
View remediation

T09 · Insecure Skill Coding Practices

Warning
Location
heartbeat/SKILL.md:98
Finding

Authenticated Heartbeat Documentation Exposes an Unencrypted WebSocket Endpoint

Content
View full analysis
Remediation
View remediation
Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Findings (17)

External Transmission

Medium
Category
Data Exfiltration
Confidence
97% confidence
Finding

This duplicate finding points to the same registration call that sends credentials to ai4trade.ai. The core risk is external transmission of sensitive user secrets across a trust boundary, which is acceptable only with informed consent and proper secret handling.

Content

Scanner excerpt · SKILL.md (reported line 116)May include surrounding context.

md
import requests

# Register Agent
response = requests.post("https://ai4trade.ai/api/claw/agents/selfRegister", json={
    "name": "MyTradingBot",
    "email": "your@email.com",
    "password": "secure_password"

External Transmission

Medium
Category
Data Exfiltration
Confidence
97% confidence
Finding

This duplicate finding points to the same registration call that sends credentials to ai4trade.ai. The core risk is external transmission of sensitive user secrets across a trust boundary, which is acceptable only with informed consent and proper secret handling.

Content

Scanner excerpt · SKILL.md (reported line 116)May include surrounding context.

md
import requests

# Register Agent
response = requests.post("https://ai4trade.ai/api/claw/agents/selfRegister", json={
    "name": "MyTradingBot",
    "email": "your@email.com",
    "password": "secure_password"

Skill Enumeration

Medium
Category
Agent Snooping
Confidence
80% confidence
Finding

Skill enumerates or reads other installed skills. Access to other skills' SKILL.md files or the skills directory reveals prompt instructions, capabilities, and secrets that should be invisible to peer skills.

Content

Scanner excerpt · SKILL.md (reported line 461)May include surrounding context.

md
For Polymarket, agents should do market discovery themselves:
- Resolve the market question and outcome by calling Polymarket public APIs directly
- Use `skills/polymarket/SKILL.md` or `https://ai4trade.ai/skill/polymarket`

Recommended publishing shape:

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 575)May include surrounding context.

bash
# Method 1: via /api/claw/agents/me
curl -H "Authorization: Bearer {token}" https://ai4trade.ai/api/claw/agents/me

# Method 2: via /api/positions
curl -H "Authorization: Bearer {token}" https://ai4trade.ai/api/positions

External Transmission

Medium
Category
Data Exfiltration
Confidence
93% confidence
Finding

The heartbeat example causes authenticated external polling to a third-party service and can expose behavioral metadata, notifications, and task contents outside the local environment. Although the file adds bounds and opt-in language, persistent or repeated polling still expands the data-sharing surface and can silently operationalize ongoing monitoring if an agent follows the example too literally.

Content

Scanner excerpt · SKILL.md (reported line 673)May include surrounding context.

md
# Bounded example. Increase max_polls only after explicit ongoing-monitoring approval.
max_polls = 1
for _ in range(max_polls):
    response = requests.post(
        "https://ai4trade.ai/api/claw/agents/heartbeat",
        headers=headers
    )

External Transmission

Medium
Category
Data Exfiltration
Confidence
98% confidence
Finding

This duplicate finding refers to the same example registration flow using a weak hard-coded sample password. The risk remains that example code can condition unsafe operator behavior and leaks credentials to a third-party service if copied directly.

Content

Scanner excerpt · SKILL.md (reported line 759)May include surrounding context.

md
import requests

# 1. Register
register_resp = requests.post("https://ai4trade.ai/api/claw/agents/selfRegister", json={
    "name": "MyBot",
    "email": "bot@example.com",
    "password": "password123"

External Transmission

Medium
Category
Data Exfiltration
Confidence
98% confidence
Finding

This duplicate finding refers to the same example registration flow using a weak hard-coded sample password. The risk remains that example code can condition unsafe operator behavior and leaks credentials to a third-party service if copied directly.

Content

Scanner excerpt · SKILL.md (reported line 759)May include surrounding context.

md
import requests

# 1. Register
register_resp = requests.post("https://ai4trade.ai/api/claw/agents/selfRegister", json={
    "name": "MyBot",
    "email": "bot@example.com",
    "password": "password123"

External Transmission

Medium
Category
Data Exfiltration
Confidence
95% confidence
Finding

This duplicate finding refers to the same authenticated strategy-publication request. The danger is unauthorized or accidental posting of content to an external service, potentially causing unwanted disclosure, account activity, or reputational harm.

Content

Scanner excerpt · SKILL.md (reported line 769)May include surrounding context.

md
headers = {"Authorization": f"Bearer {token}"}

# 2. Publish Strategy — run only after explicit approval of this exact post
strategy_resp = requests.post("https://ai4trade.ai/api/signals/strategy", headers=headers, json={
    "market": "us-stock",
    "title": "BTC Breaking Out",
    "content": "Analysis: BTC may break $100,000 this weekend...",

External Transmission

Medium
Category
Data Exfiltration
Confidence
95% confidence
Finding

This duplicate finding refers to the same authenticated strategy-publication request. The danger is unauthorized or accidental posting of content to an external service, potentially causing unwanted disclosure, account activity, or reputational harm.

Content

Scanner excerpt · SKILL.md (reported line 769)May include surrounding context.

md
headers = {"Authorization": f"Bearer {token}"}

# 2. Publish Strategy — run only after explicit approval of this exact post
strategy_resp = requests.post("https://ai4trade.ai/api/signals/strategy", headers=headers, json={
    "market": "us-stock",
    "title": "BTC Breaking Out",
    "content": "Analysis: BTC may break $100,000 this weekend...",

External Transmission

Medium
Category
Data Exfiltration
Confidence
96% confidence
Finding

This duplicate finding refers to the same follow/copy-trading mutation. In this skill's financial context, account mutations tied to signal providers can have amplified consequences compared with ordinary social actions.

Content

Scanner excerpt · SKILL.md (reported line 783)May include surrounding context.

md
print(f"Latest signals: {signals_resp.json()}")

# 4. Follow a Trader — run only after explicit approval of leader_id 10 and copy behavior
follow_resp = requests.post("https://ai4trade.ai/api/signals/follow",
    headers=headers,
    json={"leader_id": 10}
)

External Transmission

Medium
Category
Data Exfiltration
Confidence
96% confidence
Finding

This duplicate finding refers to the same follow/copy-trading mutation. In this skill's financial context, account mutations tied to signal providers can have amplified consequences compared with ordinary social actions.

Content

Scanner excerpt · SKILL.md (reported line 783)May include surrounding context.

md
print(f"Latest signals: {signals_resp.json()}")

# 4. Follow a Trader — run only after explicit approval of leader_id 10 and copy behavior
follow_resp = requests.post("https://ai4trade.ai/api/signals/follow",
    headers=headers,
    json={"leader_id": 10}
)

External Transmission

Medium
Category
Data Exfiltration
Confidence
50% 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 · copytrade/SKILL.md (reported line 68)May include surrounding context.

Register (If Not Already)

bash
POST https://api.ai4trade.ai/api/claw/agents/selfRegister
{"name": "MyFollowerBot"}

External Transmission

Medium
Category
Data Exfiltration
Confidence
50% 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 · copytrade/SKILL.md (reported line 253)May include surrounding context.

Register (If Not Already)

bash
POST https://api.ai4trade.ai/api/claw/agents/selfRegister
{"name": "MyFollowerBot"}

External Transmission

Medium
Category
Data Exfiltration
Confidence
50% 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 · tradesync/SKILL.md (reported line 70)May include surrounding context.

Register (If Not Already)

bash
POST https://api.ai4trade.ai/api/claw/agents/selfRegister
{"name": "MyFollowerBot"}

External Transmission

Medium
Category
Data Exfiltration
Confidence
50% 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 · tradesync/SKILL.md (reported line 225)May include surrounding context.

Register (If Not Already)

bash
POST https://api.ai4trade.ai/api/claw/agents/selfRegister
{"name": "MyFollowerBot"}

Description-Behavior Mismatch

Medium
Category
Not specified by scanner
Confidence
92% confidence
Finding

The manifest description limits this skill to reading heartbeat notifications with bounded polling or approved recurring monitoring. However, the file also documents broader capabilities such as fetching discussions, searching signals, retrieving replies, and checking discussion reply state, which go beyond mere heartbeat notification handling.

Content

No source excerpt is available for this finding.

Description-Behavior Mismatch

Low
Category
Not specified by scanner
Confidence
84% confidence
Finding

The manifest frames this as a read/monitor heartbeat skill, but the later heartbeat example includes a request body advertising broader agent capabilities such as "trading-signals" and "copy-trading." That broadens the apparent operational scope beyond passive notification reading.

Content

No source excerpt is available for this finding.