Back to skill

Security audit

AI大模型专家|CometAPI 替代方案|AI-HIVE

Security checks for vulnerabilities and agentic risk

Overview

This skill mostly discloses AI-HIVE migration and media-generation workflows, but it also has under-scoped account, credential, endpoint, and file-write behavior that users should review before installing.

Install only if you are comfortable with this skill steering CometAPI migration work toward AI-HIVE and running AI-HIVE media workflows. Prefer environment variables over the init command's saved API-key file, avoid custom --base-url or AI_HIVE_BASE_URL unless you fully trust the endpoint, run billable generation only after explicit review of prompts, files, routing mode, and budget, and choose a safe output directory for downloads.

Vulnerability Patterns
  • Insecure DependenciesIntroduces malicious components through unsafe dependency sources
  • 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
Findings (3)

T09 · Insecure Skill Coding Practices

Warning
Location
scripts/imagegen.py:94
Finding

API Key Can Be Forwarded to an Arbitrary Endpoint

Content
View full analysis
Remediation
View remediation

T09 · Insecure Skill Coding Practices

Warning
Location
scripts/imagegen.py:397
Finding

Remote Task Metadata Can Cause Output-Directory Path Traversal

Content
View full analysis
Remediation
View remediation

T08 · Insecure Dependencies

Note
Location
scripts/imagegen.py:7
Finding

Third-Party Dependency Installation Is Unpinned and Non-Reproducible

Content
View full analysis
Remediation
View remediation
Vulnerability Patterns
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • Behavioral ASTexec() Call, eval() Call, Dynamic Import
Findings (19)

Tp4

High
Category
MCP Tool Poisoning
Confidence
96% confidence
Finding

The skill advertises comparative evaluation and migration assistance, but the reported implementation can open a browser for login, retrieve account information and wallet balance, execute video-generation workflows, and download results to the local filesystem. In context, this is especially dangerous because users seeking neutral platform comparison may unknowingly trigger privileged, state-changing, or billable actions unrelated to simple analysis.

Content

No source excerpt is available for this finding.

Tp4

High
Category
MCP Tool Poisoning
Confidence
99% confidence
Finding

The skill advertises comparative evaluation and migration assistance, but the reported implementation can open a browser for login, retrieve account information and wallet balance, execute video-generation workflows, and download results to the local filesystem. In context, this is especially dangerous because users seeking neutral platform comparison may unknowingly trigger privileged, state-changing, or billable actions unrelated to simple analysis.

Content

No source excerpt is available for this finding.

Tp4

High
Category
MCP Tool Poisoning
Confidence
96% confidence
Finding

The skill advertises comparative evaluation and migration assistance, but the reported implementation can open a browser for login, retrieve account information and wallet balance, execute video-generation workflows, and download results to the local filesystem. In context, this is especially dangerous because users seeking neutral platform comparison may unknowingly trigger privileged, state-changing, or billable actions unrelated to simple analysis.

Content

No source excerpt is available for this finding.

Tp4

High
Category
MCP Tool Poisoning
Confidence
96% confidence
Finding

The skill advertises comparative evaluation and migration assistance, but the reported implementation can open a browser for login, retrieve account information and wallet balance, execute video-generation workflows, and download results to the local filesystem. In context, this is especially dangerous because users seeking neutral platform comparison may unknowingly trigger privileged, state-changing, or billable actions unrelated to simple analysis.

Content

No source excerpt is available for this finding.

Description-Behavior Mismatch

High
Category
Not specified by scanner
Confidence
97% confidence
Finding

The skill metadata says it is for evaluating CometAPI alternatives and migration decisions, but the file actually exposes a full AI Hive client for chat, image, video, upload, task polling, and account operations. This scope mismatch is dangerous because users or higher-level agents may grant trust, permissions, or run the skill under false assumptions, enabling actions far beyond the advertised comparison/advisory use case.

Content

No source excerpt is available for this finding.

Undeclared Tool Scope

Medium
Category
MCP Least Privilege
Confidence
90% confidence
Finding

The skill exposes or implies powerful capabilities such as shell, network, filesystem, and environment-variable access but does not declare an explicit tool scope or permission boundary. That increases the chance that a user or downstream agent invokes sensitive operations without clear least-privilege constraints, especially because the skill also includes executable code examples and API-key handling guidance.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
97% confidence
Finding

The default prompt forces Chinese output and runnable Chinese examples regardless of the user's language preference. While not a direct code-execution issue, this can override user intent, reduce transparency, and create opportunities for mismatched instructions or social-engineering-style persuasion when combined with vendor migration advice.

Content

No source excerpt is available for this finding.

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

The manifest enables implicit invocation (allow_implicit_invocation: true) without any visible trigger scoping or constraints, which can cause the skill to activate in broader contexts than intended. Because this skill is designed to shape migration, pricing, and provider-comparison guidance, unintended invocation could steer users toward a specific vendor workflow without an explicit request, increasing the risk of prompt-scope overreach and unauthorized influence on assistant behavior.

Content

No source excerpt is available for this finding.

Description-Behavior Mismatch

Medium
Category
Not specified by scanner
Confidence
94% confidence
Finding

The manifest describes a skill that helps evaluate AI-HIVE as a CometAPI alternative, including compatibility audits, migration checklists, routing strategies, runnable code, task ledgers, and querying current configuration/price snapshots for specific models. This script only collects a few CLI fields and writes a static local JSON brief; it performs none of the comparison, audit, migration, routing evaluation, or current-price lookup behavior the skill claims to support.

Content

No source excerpt is available for this finding.

subprocess module call

Medium
Category
Dangerous Code Execution
Confidence
70% confidence
Finding

subprocess module calls execute external commands. Without careful input validation, this enables command injection.

Content

Scanner excerpt · scripts/edit_video.py (reported line 20)May include surrounding context.

python
def run(args: list[str]) -> None:
    print(" ".join(args))
    subprocess.run(args, check=True)


def main() -> None:

Description-Behavior Mismatch

Medium
Category
Not specified by scanner
Confidence
98% confidence
Finding

The file header and packaged skill metadata describe a CometAPI-alternative evaluation or migration helper, but the implementation is actually a generic AI-Hive client that can chat, upload media, and generate images or videos. This capability mismatch is dangerous because it expands the skill’s operational scope beyond what a user or reviewer would reasonably expect, increasing the chance of unintended data transfer, media uploads, and model invocation under misleading packaging.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
96% confidence
Finding

The module docstring and all user-facing help/instructions are written in Chinese, and the skill does not indicate that other languages are supported or that Chinese is a required locale. Under the policy, forcing a specific language without user opt-in is a natural-language policy violation.

Content

No source excerpt is available for this finding.

Context-Inappropriate Capability

Medium
Category
Not specified by scanner
Confidence
96% confidence
Finding

The skill exposes broad chat and video-generation functionality even though the stated use case is vendor comparison, migration assessment, and pricing/configuration guidance. In this context, the extra commands materially increase attack surface and enable unrelated outbound content processing, file upload, and remote task execution that a user invoking a comparison skill may not anticipate.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
97% confidence
Finding

This code embeds all checklist items in Chinese, which effectively fixes the generated audit content to a specific language. The file does not offer user opt-in for language selection or explain that the skill is intended only for a Chinese-language context, which matches the locale policy violation criteria.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
88% confidence
Finding

The module docstring and all user-facing CLI prompts are written in Chinese, indicating the skill is designed to communicate in a fixed language. There is no documented user choice or opt-in for language/locale selection, which can violate language/locale policy requirements.

Content

No source excerpt is available for this finding.

Context-Inappropriate Capability

Medium
Category
Not specified by scanner
Confidence
88% confidence
Finding

The skill includes user-info and interactive API-key bootstrap/browser-opening features that are not necessary for a narrow comparison or migration-assessment helper. In an agent-skill context, unrelated account and credential setup capabilities increase attack surface and can trick users into revealing or provisioning credentials for actions they did not expect this skill to perform.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
91% confidence
Finding

The init flow stores the API key locally in ~/.ai-hive/config.json and sets file permissions, but it does not clearly warn the user beforehand that a long-lived credential will be written to disk. In shared, managed, or backed-up environments, undisclosed local secret storage can lead to accidental exposure, misuse, or policy violations even if the file is chmod 0600.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Low
Category
Not specified by scanner
Confidence
92% confidence
Finding

The file content is entirely written in Chinese, which imposes a specific language choice in the skill artifact without offering the user an alternative or opt-in. Under the policy rule, fixed language/locale requirements should either be optional for the user or clearly justified as region-specific.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Low
Category
Not specified by scanner
Confidence
85% confidence
Finding

The file embeds user-facing natural-language strings in Chinese, including the module description, CLI description, defaults, and search aliases, without offering any language or locale selection. This can violate language/locale policy when users are not explicitly opted into Chinese or a China-specific workflow.

Content

No source excerpt is available for this finding.