Back to skill

Security audit

1688 Bp Inquiry Evaluate

Security checks across malware telemetry and agentic risk

Overview

This skill mostly performs the promised 1688 inquiry reporting, but its Access Key setup is under-scoped and writes credentials under a different skill identity.

Review before installing. Use only where 1688 business report access is intended, avoid pasting production Access Keys into normal chat, and verify where OpenClaw stores the key. Treat generated reports as sensitive because they may include buyer IDs, inquiry summaries, and employee performance data. Static scan was clean and VirusTotal telemetry was absent, so this is Review rather than malicious.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
  • Taint TrackingDirect Taint Flow, Variable-Mediated Taint Flow, Credential Exfiltration Chain
  • MCP Least PrivilegeUnderdeclared Capability, Wildcard Permission, Missing Permission Declaration
  • MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
Findings (19)

Tainted flow: 'gateway_url' from os.environ.get (line 36, credential/environment) → requests.patch (network output)

Critical
Category
Data Flow
Content
headers = {}
        if token:
            headers["Authorization"] = f"Bearer {token}"
        resp = requests.patch(f"{gateway_url}/api/config",
                              headers=headers, json=payload, timeout=5)
        return resp.ok
    except Exception:
Confidence
90% confidence
Finding
resp = requests.patch(f"{gateway_url}/api/config", headers=headers, json=payload, timeout=5)

Lp3

Medium
Category
MCP Least Privilege
Confidence
93% confidence
Finding
The skill documentation describes access to environment variables, local files, and outbound network activity, but no explicit permission declaration or user-facing consent model is present. This creates a transparency and governance gap: a user invoking what appears to be a reporting skill may unknowingly trigger broader-capability code paths, including config reads/writes and remote communication.

Description-Behavior Mismatch

Medium
Confidence
97% confidence
Finding
The manifest presents the skill as an inquiry-quality evaluation tool, but it also performs automatic outbound telemetry on every CLI execution. That secondary behavior is unrelated to the core user task and can violate user expectations about what data is sent externally, especially when each invocation generates a remote usage report.

Context-Inappropriate Capability

Medium
Confidence
96% confidence
Finding
The telemetry reporting capability is not necessary to fulfill the stated business purpose of evaluating inquiry handling quality, so it expands the skill's operational scope beyond user expectations. Even if the payload is described as limited, unrelated outbound communication increases privacy, compliance, and trust risks because it can be changed or combined with other metadata over time.

Description-Behavior Mismatch

Medium
Confidence
96% confidence
Finding
The skill is advertised as an inquiry-evaluation/reporting capability, but this file introduces collection of an Access Key and local secret configuration for DingTalk messaging, which is outside that declared scope. Scope expansion into credential handling is dangerous because it enables the agent to solicit sensitive secrets and execute privileged local configuration actions not clearly required for the stated business-evaluation function.

Context-Inappropriate Capability

Medium
Confidence
97% confidence
Finding
The documented behavior tells the agent to ask users for an AK used for DingTalk messaging, but the skill metadata only describes evaluation and quality-reporting functions. That mismatch increases the risk of credential harvesting or unauthorized side effects because users would not reasonably expect a reporting skill to request and install messaging credentials.

Context-Inappropriate Capability

Medium
Confidence
96% confidence
Finding
The skill's stated purpose is inquiry quality evaluation and report generation, but this code adds undisclosed outbound telemetry on every CLI run. Even though the payload is limited to skill metadata, the hidden network behavior expands the trust boundary and creates privacy, compliance, and supply-chain risk because users are not clearly informed or given control.

Description-Behavior Mismatch

Medium
Confidence
97% confidence
Finding
The manifest describes an evaluation/reporting skill, but the implementation performs an outbound POST to a gateway endpoint unrelated to generating the user-requested report. This mismatch is a real security concern because hidden network actions can surprise users and enable unauthorized data collection or policy violations.

Description-Behavior Mismatch

High
Confidence
97% confidence
Finding
This module persists API keys into global skill configuration even though the declared skill purpose is inquiry-quality evaluation/reporting. That hidden credential-management capability expands the trust boundary and can enable unauthorized secret collection, storage, and later use, especially because users would not expect a reporting skill to modify platform-wide config.

Context-Inappropriate Capability

High
Confidence
96% confidence
Finding
The skill implements AK validation, retrieval, transmission, and persistence capabilities that are unrelated to its advertised analytics/reporting function. This creates hidden functionality that can collect and manage credentials under misleading pretenses, increasing the likelihood of abuse or unsafe deployment.

Intent-Code Divergence

Medium
Confidence
88% confidence
Finding
The module docstring explicitly says this is an AK configuration service, which contradicts the manifest describing an inquiry-evaluation/reporting skill. Such capability mismatch is a supply-chain and trust issue: reviewers or users may enable the skill for analytics while it actually handles secrets and system configuration.

Missing User Warnings

Medium
Confidence
95% confidence
Finding
Although the markdown mentions usage reporting, it does not clearly warn users that every CLI invocation automatically sends telemetry to a remote gateway. This weak disclosure is dangerous because users may reasonably believe they are only running a local read-only evaluation workflow, while metadata about their activity is being transmitted externally without explicit notice or consent.

Missing User Warnings

Medium
Confidence
92% confidence
Finding
The CLI unconditionally invokes telemetry reporting after every command, and this file shows no user notice, consent flow, or opt-out mechanism. Even if the telemetry is intended for benign usage analytics, silent collection can expose command usage patterns and potentially sensitive operational metadata, which is especially concerning for a business-evaluation skill that may process merchant or salesperson-related data.

Missing User Warnings

Medium
Confidence
89% confidence
Finding
The skill explicitly documents returning detailed inquiry records containing buyer identifiers, chat transcripts, item links, and salesperson performance data, but provides no privacy guardrails, minimization guidance, or authorization checks. In an agent setting, this can lead to overexposure of personal and commercially sensitive conversation data to users who may not need full transcript-level access.

Missing User Warnings

Medium
Confidence
90% confidence
Finding
The skill instructs the agent to invoke a CLI that uses a configured AK to query an external service, but it does not disclose that credentials will be used or that user-supplied date-range data will be transmitted externally. In an agent setting, this can cause unintended credential use and opaque outbound data access, reducing informed consent and making unauthorized or unexpected data retrieval more likely.

Missing User Warnings

Medium
Confidence
98% confidence
Finding
The skill instructs the agent to extract the user's Access Key from chat and pass it directly to a CLI command, with no warning about the sensitivity of the credential or the risks of exposing it via command-line arguments. This is dangerous because secrets entered in chat and process arguments may be logged, retained in histories, or exposed to other local users and tooling, leading to credential compromise.

Missing User Warnings

Medium
Confidence
98% confidence
Finding
Telemetry is sent on every CLI invocation, and failures are silently ignored, which makes the behavior hard for users to notice or audit. Repeated undisclosed outbound calls increase privacy and governance risk, especially in enterprise environments where unexpected network access may violate policy.

Missing User Warnings

Medium
Confidence
83% confidence
Finding
The file-write path stores the API key into persistent global configuration without any visible user-facing notice in this code path. Secret persistence increases exposure through local file compromise, backups, accidental sharing, and later reuse by other components, and it is more concerning because the skill's stated purpose does not imply credential storage.

Missing User Warnings

Medium
Confidence
84% confidence
Finding
The network path sends the API key to a gateway service without any user-visible disclosure in this module. Transmitting secrets off-process or off-host changes the risk profile substantially, and undisclosed transfer is especially problematic for a skill that presents itself as evaluation/reporting rather than credential setup.

VirusTotal

VirusTotal findings are pending for this skill version.

View on VirusTotal

Static analysis

No suspicious patterns detected.