Back to skill

Security audit

1Panel API Skills

Security checks for vulnerabilities and agentic risk

Overview

This is a documentation-only 1Panel server-management skill, but it guides agents through broad administrative and secret-handling operations with weak safety and transport guidance.

Install only if you intend to let the agent help operate a trusted 1Panel server. Use HTTPS/WSS only, treat the 1Panel API key and all generated tokens as administrator credentials, avoid pasting production secrets unless necessary, and require explicit confirmation before destructive or host-level actions such as deleting files/databases, changing SSH/firewall/Docker/Nginx config, partitioning disks, running scripts, or creating persistent services.

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 (2)

T09 · Insecure Skill Coding Practices

Error
Location
README.md:108
Finding

Administrative API Authentication Tokens Transmitted Over Plaintext HTTP

Content
View full analysis
Remediation
View remediation

T09 · Insecure Skill Coding Practices

Error
Location
docs/SKILL-process.md:225
Finding

Process Monitoring and Termination Examples Use Unencrypted and Incompletely Authenticated Connections

Content
View full analysis
{ ws.send(JSON.stringify({ type: 'ps', pid: 0, name: '', username: '' })); }; ws.onmessage = (event) => { const processes = JSON.parse(event.data); console.log(processes); }; ``` ```bash curl -X POST http://your-1panel-host:65432/api/v1/process/stop \ -H "Content-Type: application/json" \ -d '{"PID": 1234}' ``` ```bash curl -X POST http://your-1panel-host:65432/api/v1/process/listening ``` ```bash curl http://your-1panel-host:65432/api/v1/process/1234 ``` ### Technical Analysis The process-management examples use plaintext `ws://` and `http://` connections. The examples also omit all authentication headers, cookies, or other authorization controls. The process WebSocket response documented in this file includes process command lines, environment variables, open files, and network connections. Environment variables commonly contain database passwords, cloud credentials, API tokens, service endpoints, and other secrets. Sending this information over an unencrypted WebSocket exposes it to passive network interception. The `/process/stop` endpoint performs a privileged and destructive operation. Plaintext HTTP provides no integrity protection, so on-path traffic can be observed or altered. The absence of authentication in the examples creates two possible unsafe outcomes: - If the server requires authentication, the documentation is incomplete and may cause users to disable or bypass security controls while troubleshooting. - If the endpoint accepts the documented unauthenticated request, any network client capable of reaching it could terminate processes or retrieve sensitive process information. The actual serv ...[truncated 2180 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
  • MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
Findings (124)

Missing User Warnings

High
Category
Not specified by scanner
Confidence
95% confidence
Finding

This section documents submission of highly sensitive values such as API keys and agent tokens with no guidance on secure handling, storage, masking, or transport requirements. In an agent-skill context, normalizing raw secret collection increases the chance that downstream tools, logs, prompts, or operators will expose credentials, leading to account takeover or abuse of third-party AI services.

Content

No source excerpt is available for this finding.

Missing User Warnings

High
Category
Not specified by scanner
Confidence
96% confidence
Finding

These endpoints ingest API keys, app secrets, and bot tokens for external providers and messaging platforms, yet the documentation provides no safeguards for secret lifecycle management. Because this skill targets agent and integration setup, the absence of secure-handling guidance materially raises the likelihood of credential leakage through logs, prompts, screenshots, backups, or improperly protected storage.

Content

No source excerpt is available for this finding.

Missing User Warnings

High
Category
Not specified by scanner
Confidence
96% confidence
Finding

The skill documents an endpoint that accepts raw email credentials, including a password, for configuration testing, yet provides no warning about secret handling, transport security, or logging exposure. In an agent-skill context, this is riskier because operators may paste production credentials into tools that could log requests, display them in traces, or send them over insecure HTTP as shown in the base URL example.

Content

No source excerpt is available for this finding.

Missing User Warnings

High
Category
Not specified by scanner
Confidence
97% confidence
Finding

This endpoint documents deleting scan rules and includes a 'removeInfected' option that may delete infected files, a destructive and potentially irreversible action affecting user data. No warning is provided about file deletion consequences or the need for confirmation.

Content

No source excerpt is available for this finding.

Missing User Warnings

High
Category
Not specified by scanner
Confidence
95% confidence
Finding

Replacing the complete Fail2ban configuration is a high-impact operation that can break the service, weaken protections, or cause administrative lockout if invalid. The markdown does not include any explicit caution about backup, validation, or rollback.

Content

No source excerpt is available for this finding.

Credential Access

High
Category
Privilege Escalation
Confidence
60% confidence
Finding

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

Content

Scanner excerpt · docs/SKILL-apps.md (reported line 38)May include surrounding context.

md
return path.Join(i.GetAppPath(), i.Name, "docker-compose.yml")
}

// 获取 .env 文件路径
func (i *AppInstall) GetEnvPath() string {
    return path.Join(i.GetAppPath(), i.Name, ".env")
}

Credential Access

High
Category
Privilege Escalation
Confidence
60% confidence
Finding

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

Content

Scanner excerpt · docs/SKILL-containers.md (reported line 605)May include surrounding context.

md
return path.Join(i.GetAppPath(), i.Name, "docker-compose.yml")
}

// 获取 .env 文件路径
func (i *AppInstall) GetEnvPath() string {
    return path.Join(i.GetAppPath(), i.Name, ".env")
}

Credential Access

High
Category
Privilege Escalation
Confidence
60% confidence
Finding

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

Content

Scanner excerpt · docs/SKILL-apps.md (reported line 40)May include surrounding context.

// 获取 .env 文件路径 func (i *AppInstall) GetEnvPath() string { return path.Join(i.GetAppPath(), i.Name, ".env") }

text

Context-Inappropriate Capability

High
Category
Not specified by scanner
Confidence
94% confidence
Finding

The /files/mount endpoint exposes system-level mount functionality inside a file-management skill, which is substantially more privileged than ordinary file browsing or editing. Mount operations can alter visibility of devices and filesystems, expose sensitive host data, or be abused to access or modify resources outside the expected file-management scope.

Content

No source excerpt is available for this finding.

Missing User Warnings

High
Category
Not specified by scanner
Confidence
97% confidence
Finding

The documented disk partition endpoint performs an inherently destructive operation that can erase data, break filesystems, or render a host unbootable, yet the skill provides no warning or guardrail language. In an agent workflow, omission of irreversible-impact warnings makes accidental invocation substantially more likely and more dangerous.

Content

No source excerpt is available for this finding.

Missing User Warnings

High
Category
Not specified by scanner
Confidence
95% confidence
Finding

Firewall rule changes and firewall start/stop/rebind operations can immediately sever connectivity, expose services, or lock administrators out of the host. Because the skill omits warnings about network disruption and remote-access loss, an agent or operator may apply dangerous rule changes without understanding the blast radius.

Content

No source excerpt is available for this finding.

Missing User Warnings

High
Category
Not specified by scanner
Confidence
97% confidence
Finding

These SSH secret endpoints explicitly handle private keys and passphrases, and the search response even includes passPhrase, publicKey, and privateKey fields. Exposing sensitive credential material in a broadly consumable skill without strong warnings or minimization guidance creates serious risk of credential theft, lateral movement, and unauthorized host access.

Content

No source excerpt is available for this finding.

Missing User Warnings

High
Category
Not specified by scanner
Confidence
95% confidence
Finding

Deleting SSH secrets can permanently remove credentials required for automation or emergency access, potentially causing outages or administrative lockout. The absence of any warning about irreversibility or dependency impact makes accidental destructive use more plausible.

Content

No source excerpt is available for this finding.

Missing User Warnings

High
Category
Not specified by scanner
Confidence
96% confidence
Finding

The SSH configuration and service-management endpoints can weaken authentication, enable root login, change listen addresses/ports, or restart/stop SSH entirely, all of which can either reduce host security or cut off remote administration. Without warnings and guardrails, these are highly dangerous capabilities to expose through an agent-oriented skill.

Content

No source excerpt is available for this finding.

Missing User Warnings

High
Category
Not specified by scanner
Confidence
96% confidence
Finding

The tool and Supervisor endpoints can edit service configs, start/stop services, and define process commands and environments, effectively enabling arbitrary command execution under service-managed contexts. Documenting these privileged controls without warnings about disruption and code-execution risk makes the skill especially dangerous in an automated agent environment.

Content

No source excerpt is available for this finding.

Missing User Warnings

High
Category
Not specified by scanner
Confidence
95% confidence
Finding

This endpoint allows raw replacement of the website Nginx configuration file via an arbitrary content field, which is significantly more dangerous than scoped parameter updates. Without warnings or safeguards, an agent could introduce invalid config, disable protections, expose files, or proxy traffic unsafely, causing outage or security compromise.

Content

No source excerpt is available for this finding.

Missing User Warnings

High
Category
Not specified by scanner
Confidence
96% confidence
Finding

The documented API includes parameters to execute arbitrary custom shell content (execShell and shell) during certificate operations. In the context of a panel/API skill for infrastructure management, exposing shell execution is highly dangerous because it can enable remote command execution, persistence, or full host compromise if callers can supply or influence these fields.

Content

No source excerpt is available for this finding.

Context-Inappropriate Capability

High
Category
Not specified by scanner
Confidence
96% confidence
Finding

The API explicitly supports executing arbitrary custom shell script content through the SSL management workflow via execShell and shell. In an admin-facing panel this creates a direct command-execution capability that can be abused for remote code execution if the endpoint is exposed to untrusted users, invoked by an agent, or insufficiently permission-gated.

Content

No source excerpt is available for this finding.

Missing User Warnings

High
Category
Not specified by scanner
Confidence
91% confidence
Finding

The markdown describes uploading raw private key material and later shows privateKey in the response schema, but it provides no warning that these fields contain highly sensitive secrets. Documentation should clearly disclose the sensitivity of private keys and caution users about storage, transmission, and logging risks.

Content

No source excerpt is available for this finding.

Context-Inappropriate Capability

High
Category
Not specified by scanner
Confidence
99% confidence
Finding

The documented normal SSL detail response includes raw privateKey material, meaning any caller able to fetch certificate details can obtain the secret key. Exposure of private keys fully compromises TLS identity, enabling impersonation, traffic interception, and long-term credential theft until the certificate is revoked and replaced.

Content

No source excerpt is available for this finding.

Missing User Warnings

High
Category
Not specified by scanner
Confidence
96% confidence
Finding

The response example publishes raw private key material in normal API output and omits any warning that the endpoint may return credential secrets. In this skill context, documentation is actionable for agents and operators, so demonstrating secret-bearing responses increases the chance of unsafe integration, logging, or automated exfiltration of key material.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
90% confidence
Finding

This markdown file instructs users to configure a panel address and API key, and later shows authenticated API usage including creating websites, but it does not warn that the API key is sensitive or that authenticated calls can modify server configuration and data. For a server-management skill covering backups, files, databases, scripts, and websites, the description should disclose privacy and system-integrity implications before use.

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 108)May include surrounding context.

bash
# 获取容器列表
curl -X GET "http://192.168.1.100:8888/api/v2/containers/list" \
  -H "1Panel-Token: your_token" \
  -H "1Panel-Timestamp: 1709548800" \
  -H "Content-Type: application/json"

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
90% confidence
Finding

The skill asks the user to provide a 1Panel API key for broad server-management operations but does not warn that this credential is highly sensitive or that misuse can affect websites, containers, databases, files, backups, and system settings. In an agent context, collecting such a credential without explicit scoping, least-privilege guidance, storage/handling warnings, or confirmation for destructive actions increases the risk of credential exposure and unauthorized administrative changes.

Content

No source excerpt is available for this finding.

Description-Behavior Mismatch

Medium
Category
Not specified by scanner
Confidence
89% confidence
Finding

The manifest describes this skill as providing API interface documentation for 1Panel modules, but this file documents endpoints that create, update, start, stop, and delete runtime resources such as MCP servers, TensorRT containers, Ollama models, and Agents. Those are active administrative and orchestration operations, not merely passive documentation or limited read-oriented lookup behavior.

Content

No source excerpt is available for this finding.