Back to skill

Security audit

Enterprise Bid Document AI — 企业招投标文书AI助手

Security checks for vulnerabilities and agentic risk

Overview

The verdict is benign with medium confidence: scanners found no AIG issues and judged the reported concerns largely consistent with the skill’s disclosed procurement-assistance purpose, though its installation instructions use an unpinned CLI and one reference was unresolved.

Before installing, consider that the CLI version is not pinned, so a future installation could retrieve a different release. Review package provenance and integrity and, where supported, pin reviewed CLI and skill versions. Skillspector scanned the listed components, but coverage is marked partial because a local reference in SKILL.md could not be resolved; the advertised bidding functionality therefore could not be verified from the supplied chunk.

Vulnerability Patterns
  • 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
  • Unauthorized Access and Privilege EscalationObtains permissions beyond the task's legitimate needs
Vulnerability Patterns
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
  • 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
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
Findings (9)

Tp2

High
Category
MCP Tool Poisoning
Confidence
85% confidence
Finding

Mixing characters from multiple Unicode scripts in a single identifier is a common technique to create visually ambiguous tool names.

Content

No source excerpt is available for this finding.

Description-Behavior Mismatch

High
Category
MCP Tool Poisoning
Confidence
96% confidence
Finding

The supplied code is an installation command, not an implementation of the declared bidding assistant. Installing executable skill content is materially different from processing procurement documents and is not disclosed in the description. The installed skill might provide the advertised features, but its contents and behavior cannot be verified from this chunk.

Content

No source excerpt is available for this finding.

MCP server referenced without pinned version: 'npx clawhub'.

Medium
Category
MCP Rug Pull
Confidence
93% confidence
Finding

The installation example executes an npm CLI without pinning its version, allowing the code retrieved on a fresh installation to change over time. Although this is not an MCP server reference, a compromised or malicious package release could execute with the installing user's permissions and potentially access confidential bidding documents.

Content

No source excerpt is available for this finding.

Autonomous Decision Making

Medium
Category
Excessive Agency
Confidence
80% confidence
Finding

Skill enables autonomous high-impact decisions without human-in-the-loop verification. Critical operations (destructive commands, financial transactions, data deletion) should require explicit user confirmation.

Content

Scanner excerpt · README.md (reported line 57)May include surrounding context.

md
- **Bid documents are confidential.** Uploaded files may contain pricing, technical proposals, and business-sensitive content.
- **Sanitize before upload.** Remove or redact sensitive data (specific prices, client names, core technical parameters) before submitting any document for analysis.
- **No guaranteed deletion.** Uploaded documents are not guaranteed to be automatically deleted or anonymized after processing.
- **Your responsibility.** Ensure your organization's data security policy allows using third-party AI tools for business document analysis.

## 🌐 GitHub

Autonomous Decision Making

Medium
Category
Excessive Agency
Confidence
80% confidence
Finding

Skill enables autonomous high-impact decisions without human-in-the-loop verification. Critical operations (destructive commands, financial transactions, data deletion) should require explicit user confirmation.

Content

Scanner excerpt · SKILL.md (reported line 626)May include surrounding context.

md
- **Bid documents are confidential.** Uploaded files may contain pricing, technical proposals, and business-sensitive content.
- **Sanitize before upload.** Remove or redact sensitive data (specific prices, client names, core technical parameters, personnel identity information) before submitting documents for analysis.
- **No guaranteed deletion.** Uploaded documents are not guaranteed to be automatically deleted or anonymized after processing.
- **Your responsibility.** You are responsible for ensuring your organization's data security policies allow using third-party AI tools for business document analysis.

### Quick Start

Autonomous Decision Making

Medium
Category
Excessive Agency
Confidence
80% confidence
Finding

Skill enables autonomous high-impact decisions without human-in-the-loop verification. Critical operations (destructive commands, financial transactions, data deletion) should require explicit user confirmation.

Content

Scanner excerpt · SKILL.md (reported line 653)May include surrounding context.

md
- **Sanitize before use:** Remove or redact sensitive data (specific prices, client names, core technical parameters, personnel identity information) before submitting any document for analysis.
- **Authorization:** Ensure you have permission to use the relevant document content for AI-assisted analysis.
- **Compliance:** Confirm your organization's data security policy allows using third-party AI tools for business document analysis.
- **No guaranteed deletion:** Uploaded documents are not guaranteed to be automatically deleted or anonymized after processing.

**What this skill does NOT do:**
- Does NOT facilitate bid-rigging, collusion, or any form of corruption.

MCP server referenced without pinned version: 'npx clawhub'.

Medium
Category
MCP Rug Pull
Confidence
93% confidence
Finding

The repeated installation command invokes an unpinned npm CLI, so a fresh installation may execute a different release without review. This is a conditional supply-chain exposure rather than evidence of malicious code or an MCP server, but execution under an enterprise user's account could expose sensitive business files.

Content

No source excerpt is available for this finding.

MCP server referenced without pinned version: 'npx clawhub'.

Medium
Category
MCP Rug Pull
Confidence
93% confidence
Finding

The Quick Start command invokes clawhub through npx without specifying a reviewed version. Although it references a CLI rather than an MCP server, a compromised package release could execute with local user permissions, creating a supply-chain risk in an environment that may hold confidential bids.

Content

No source excerpt is available for this finding.

MCP server referenced without pinned version: 'npx clawhub'.

Medium
Category
MCP Rug Pull
Confidence
93% confidence
Finding

The inline installation command leaves the executable npm package version unspecified, making installations dependent on a mutable package release. Exploitation would require a malicious or compromised release, but the resulting local execution could affect business documents accessible to the installer.

Content

No source excerpt is available for this finding.