Back to skill

Security audit

competitor-monitor-weekly

Security checks for vulnerabilities and agentic risk

Overview

The skill’s competitor-report workflow is mostly transparent, but its setup relies on mutable external installs and its automated report forwarding needs careful review before use.

Review and pin the installer and dependent skill versions before installing. Confirm the Feishu, DingTalk, email, and n8n destinations are approved for competitor intelligence, and consider adding a human review step before scheduled reports are sent automatically.

Vulnerability Patterns
  • Insecure DependenciesIntroduces malicious components through unsafe dependency sources
  • 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)

T08 · Insecure Dependencies

Warning
Location
SKILL.md:148
Finding
Unpinned Third-Party Dependency Installation## Vulnerability Details **File Location**: `SKILL.md`, lines 148-152 **Vulnerability Type**: Unpinned third-party dependencies and mutable package installer **Risk Level**: Medium ### Vulnerable Code ```bash npx clawhub@latest install agent-reach npx clawhub@latest install brave-search npx clawhub@latest install summarize npx clawhub@latest install paddleocr-doc-parsing npx clawhub@latest install n8n-workflow-automation ``` ### Technical Analysis The documented installation process invokes the mutable `latest` release of `clawhub` and installs five remotely sourced Skills without fixed versions or integrity verification. Consequently, the commands may retrieve code that differs from the code available when this project was audited. Because `npx` can download and execute the selected package, compromise of the package registry, publisher account, installer package, or any referenced Skill release could turn this installation procedure into a supply-chain execution channel. No malicious payload is present in the audited project itself; the risk arises from trusting mutable, unverified external dependencies. ### Attack Path 1. An attacker compromises a relevant package publisher, registry entry, distribution account, or future dependency release. 2. The attacker publishes malicious code under the release selected by `@latest` or within one of the unpinned Skills. 3. A user follows the installation commands in `SKILL.md`. 4. `npx` retrieves and executes the mutable `clawhub` package. 5. The installer retrieves and activates the compromised Skill dependency. 6. Malicious code executes with the permissions of the user or automation environment running the installation. ### Impact Assessment Successful exploitation could permit arbitrary code execution with the installing user's privileges. Depending on that environment, an attacker could access local files and credentials available to the user, alter the OpenClaw workspace, tamper with generated reports, ac ...[truncated 240 chars]
Remediation
## Remediation Suggestions 1. Pin `clawhub` to a specific reviewed version rather than using `@latest`. 2. Pin every installed Skill to an explicit, reviewed version or immutable revision. 3. Use lockfiles and cryptographic integrity hashes where supported. 4. Restrict installation to an approved registry and validate package provenance or signatures. 5. Audit the complete transitive dependency set before deployment. 6. Perform installation in a sandboxed, least-privilege environment without production secrets. 7. Introduce an update-review process so dependency changes are tested and approved before rollout.
Vulnerability Patterns
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
Findings (11)

Missing User Warnings

Medium
Confidence
95% confidence
Finding
The skill description does not clearly warn that it will automatically collect external content on a schedule and distribute generated reports to external systems. This weakens informed consent and can lead users to enable automated data collection and sharing without understanding privacy, compliance, or data-leakage implications.

Missing User Warnings

Medium
Confidence
94% confidence
Finding
The configuration enables report delivery to webhook and email destinations without any warning that collected or summarized content will leave the local environment. If sensitive competitor intelligence, internal annotations, or misclassified content are included, this can result in unintended disclosure to third-party services or wrong recipients.

Rp1

Medium
Category
MCP Rug Pull
Confidence
92% confidence
Finding
The installation command uses `npx clawhub@latest`, which pulls and executes whatever package version is current at runtime. This creates a supply-chain risk: a compromised upstream package, typosquatted dependency, or malicious newly published version could execute arbitrary code on the user's machine during install.

Rp1

Medium
Category
MCP Rug Pull
Confidence
92% confidence
Finding
This command again invokes `npx clawhub@latest`, causing remote code execution from an unpinned package version. Repeating the pattern across multiple install steps increases exposure because every invocation trusts the latest published package state at execution time.

Rp1

Medium
Category
MCP Rug Pull
Confidence
92% confidence
Finding
Using `npx` with an unpinned `@latest` package delegates execution to mutable third-party code. In a skill installation guide, this is especially risky because users may copy-paste commands directly, turning documentation into a practical code-execution vector if the package supply chain is compromised.

Rp1

Medium
Category
MCP Rug Pull
Confidence
92% confidence
Finding
The command references `npx clawhub@latest` without version pinning, exposing users to arbitrary behavior changes and potential malicious package updates. Because this skill is auto-generated and intended for automation workflows, users may trust and automate these instructions, amplifying the blast radius of a compromised package.

Rp1

Medium
Category
MCP Rug Pull
Confidence
92% confidence
Finding
This final `npx clawhub@latest` invocation carries the same supply-chain risk as the others: execution of mutable remote code at install time. The repeated pattern indicates insecure installation guidance rather than an isolated mistake, making accidental unsafe adoption more likely.

Natural-Language Policy Violations

Medium
Confidence
95% confidence
Finding
This manifest hard-codes the output language to "zh-CN", which is a natural-language locale constraint. The file does not indicate that users can opt into another language or that the restriction is required for a region-specific compliance purpose.

Natural-Language Policy Violations

Medium
Confidence
97% confidence
Finding
The template content is entirely written in Chinese, including headings, labels, and footer text, which indicates the skill output is fixed to a specific language. Under the policy, locale or language constraints should either offer user opt-in or be clearly documented as justified for a region-specific use case, which is not present in this file.

Missing User Warnings

Medium
Confidence
91% confidence
Finding
The workflow automatically forwards generated reports and configured channels to an external automation skill for downstream distribution, but the manifest provides no disclosure, approval gate, recipient constraints, or content-sensitivity checks. Because the collected and analyzed content may include proprietary intelligence, accidental inclusion of sensitive data could be propagated to external systems or broad channels without user awareness.

Natural-Language Policy Violations

Low
Confidence
86% confidence
Finding
The manifest description is written only in Chinese, which can impose a language expectation on users without any opt-in or documented locale constraint. Under the language/locale policy, natural-language content should not force a specific language unless the skill clearly offers a choice or justifies the regional scope.

Static analysis

No suspicious patterns detected.