Back to skill

Security audit

PayAClaw 日报提交助手

Security checks for vulnerabilities and agentic risk

Overview

The skill fits a PayAClaw/OpenClawLog publishing workflow, but it needs Review because it handles credentials and public posts with weak safeguards and adds undisclosed promotional text to generated reports.

Install only if you are comfortable giving the skill credentials for PayAClaw and OpenClawLog and publishing report content externally. Use dedicated low-privilege/app-specific credentials, keep secrets in a secure store, verify the XML-RPC URL is exactly the intended OpenClawLog endpoint, preview/redact reports before publishing, and remove the promotional footer if it is not wanted.

Vulnerability Patterns
  • Skill Instruction HijackingAlters the agent's session goals or safety constraints when the skill loads
  • Insecure DependenciesIntroduces malicious components through unsafe dependency sources
  • Insecure Skill Coding PracticesFinds exploitable flaws such as hardcoded secrets or command injection
  • 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
Findings (4)

T01 · Skill Instruction Hijacking

Error
Location
scripts/publish_report.py:91
Finding

Undisclosed Promotional Content Injected into Generated Reports

Content
View full analysis

OpenClaw 赚钱平台:https://payaclaw.com/{end_note}

``` ### Technical Analysis The default report generator unconditionally adds a PayAClaw promotional footer to every generated report. This content is unrelated to the structured report sections and cannot be disabled through a documented option. The Skill documentation presents the generator as a way to create a work report containing accomplishments, problems, plans, and insights. It does not clearly disclose that the generated report will always contain third-party promotional content. Consequently, the generated output is altered beyond the user's stated report content and may subsequently be published publicly. This behavior best matches instruction or output hijacking because normal report-generation requests are repurposed to distribute promotional content. ### Attack Path 1. A user invokes the script with the `--generate` option or otherwise calls `generate_default_report()`. 2. The function builds the requested report from the supplied report data. 3. The hardcoded PayAClaw promotional footer is appended without an explicit opt-in. 4. The generated HTML is saved, copied, or published to a public platform. 5. The user's report therefore distributes third-party promotional content without clear informed consent. ### Impact Assessment The issue does not grant operating-system privileges or direct access to user accounts. Its scope is the integrity of generated and published content. Potential consequences include: - Unauthorized modification of user-facing reports. - Unintentional publication of advertising or promotional material. - Reputational damage to users or organizations publishing the report. - Policy, disclosure, or content-compliance violati ...[truncated 36 chars]
Remediation
View remediation

T09 · Insecure Skill Coding Practices

Error
Location
scripts/publish_report.py:99
Finding

WordPress Credentials Can Be Transmitted to an Arbitrary XML-RPC Endpoint

Content
View full analysis
Remediation
View remediation

T09 · Insecure Skill Coding Practices

Warning
Location
scripts/publish_report.py:99
Finding

Password Accepted and Documented as a Command-Line Argument

Content
View full analysis
--password --title --content <content> ``` The argument parser explicitly accepts the password this way: ```python parser.add_argument('--password', help='WordPress password') ``` ### Technical Analysis Command-line arguments are generally unsuitable for secrets. Depending on the operating system, shell, process supervisor, and logging environment, command-line values can be exposed through: - Shell history files. - Process-listing tools. - Process accounting. - CI/CD job logs. - Terminal-session recordings. - Monitoring and telemetry systems. - Error reports or command tracing. Because the script's own usage documentation recommends this invocation pattern, users are likely to expose credentials even when operating the tool as intended. ### Attack Path 1. The user follows the documented usage and supplies the WordPress password with `--password`. 2. The shell records the complete command in its history, or the operating system exposes the argument through process inspection. 3. A local user, administrator, monitoring agent, or log reader accesses the recorded command. 4. The observer recovers the plaintext WordPress password. 5. The recovered password is used to authenticate to OpenClawLog or another service where the credential was reused. ### Impact Assessment Exploitation requires access to process information, command history, logs, or telemetry produced by the execution environment. No additional system privilege is directly granted by the vulnerable script. If exposed, the password may enable: - Unauthorized access to the associated WordPress account. - ...[truncated 220 chars]
Remediation
View remediation

T08 · Insecure Dependencies

Note
Location
package.json:31
Finding

Python Dependencies Are Unpinned and Lack Integrity Verification

Content
View full analysis
Remediation
View remediation
Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • MCP Least PrivilegeUnderdeclared Capability, Wildcard Permission, Missing Permission Declaration
  • MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
Findings (9)

Tp4

High
Category
MCP Tool Poisoning
Confidence
92% confidence
Finding

The declared description emphasizes a broader PayAClaw automation workflow: submitting PayAClaw日报 tasks, posting to OpenClawLog, and submitting solutions to PayAClaw to earn points, possibly in batch. The supplied code only does two limited things: generate a canned HTML日报 template and publish provided content to OpenClawLog through WordPress XML-RPC. There is no code for logging into PayAClaw, calling any PayAClaw endpoint, submitting tasks/solutions, earning积分, or batch processing tasks. While the OpenClawLog publishing portion aligns partially with the description, the core PayAClaw submission capability is absent, making the description materially broader than the actual behavior.

Content

No source excerpt is available for this finding.

Undeclared Tool Scope

Medium
Category
MCP Least Privilege
Confidence
82% confidence
Finding

The skill describes actions that would read and write data, store returned secrets, and invoke local scripts, but it does not declare an explicit tool scope or permissions boundary. In an agent environment, missing scope declarations can cause overly broad or ambiguous tool access, increasing the chance of unintended file access or unsafe execution paths.

Content

No source excerpt is available for this finding.

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

The manifest description lists generic triggers such as "提交日报", "工作日报", and "自动日报" without narrowing context or providing exclusions. These phrases could match many ordinary user requests about writing or submitting a daily report, causing unintended invocation outside the specific PayAClaw/OpenClawLog workflow.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

The skill explicitly instructs users to obtain, save, and use API keys plus WordPress credentials, and to publish content publicly, without any warning about secret handling, least privilege, or disclosure risks. That creates a realistic chance of credential leakage, unsafe storage, accidental reuse in insecure contexts, and unintended public exposure of sensitive content.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
92% confidence
Finding

The manifest requires highly sensitive credentials (a WordPress username/password and an API key) but provides no warning about safe storage, least-privilege use, or the risks of entering account passwords into an automation skill. In this skill’s context, that is more dangerous because it automates publishing and earning workflows against third-party services, so compromised secrets could enable account takeover, unauthorized posting, or abuse of the PayAClaw account.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
91% confidence
Finding

The manifest advertises automated generation and submission of work reports to external services while requesting read-write filesystem and network API access, yet it does not provide any user-facing disclosure about what local data may be read, what content will be transmitted, or that remote publication/submission will occur. In a skill that can publish logs and submit earning-related tasks, this increases the risk of unintended data exfiltration, accidental posting of sensitive content, or surprising external side effects if a user invokes it without clear informed consent.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
88% confidence
Finding

The file mixes English CLI usage with Chinese-only descriptions and generated default report content, including the report title and section text, without offering a language selection or documenting that the skill is intentionally region-specific. This can violate language/locale policy when a skill forces a specific language without user opt-in.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
93% confidence
Finding

This code reads username and password from a credentials file and then uses them to authenticate to a remote XML-RPC endpoint for publishing. Although the script's purpose is publishing, there is no visible warning or disclosure in the CLI help, comments, or runtime output that credentials and report content will be sent to an external service.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Low
Category
Not specified by scanner
Confidence
80% confidence
Finding

The display name, description, summary, and setup descriptions are presented only in Chinese, which can impose a language constraint on users without any opt-in or documented regional justification. This is a natural-language policy concern because the manifest does not offer an alternative language or explain that the skill is intended only for Chinese-speaking users.

Content

No source excerpt is available for this finding.