Back to skill

Security audit

Internet Archive Skill

Security checks for vulnerabilities and agentic risk

Overview

This Internet Archive skill is coherent, but it needs Review because it can install unpinned software and make public/account-changing uploads or metadata edits without built-in confirmations.

Install only if you are comfortable with an agent using Internet Archive credentials and making remote changes on your account. Review every upload path, identifier, collection, and metadata change before running it; avoid using it on shared machines unless ia credentials are protected; and prefer manually installing a reviewed, pinned internetarchive package instead of using the install intent.

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
internet-archive.py:64
Finding

Unpinned Third-Party Package Installation

Content
View full analysis

Vulnerability Details

File Location: internet-archive.py:64-76
Vulnerability Type: Supply-chain risk from unpinned dependency installation
Risk Level: Medium

Vulnerable Code

python
def install_ia() -> Dict[str, Any]:
    """Install the ia CLI using uv, pipx, or pip."""
    methods = [
        ['uv', 'tool', 'install', 'internetarchive'],
        ['pipx', 'install', 'internetarchive'],
        ['pip', 'install', 'internetarchive']
    ]
    
    for method in methods:
        try:
            result = subprocess.run(method, capture_output=True, text=True, timeout=120)
            if result.returncode == 0:
                return {

Technical Analysis

The install intent invokes uv, pipx, or pip to install the latest version of the internetarchive package. It does not pin an audited version, verify package hashes or signatures, or enforce a trusted package index.

Package resolution may therefore change after the skill has been reviewed. These package managers may also honor user- or system-level index configuration, including environment variables and configuration files that redirect resolution to a custom mirror. Although subprocess arguments are passed as an array and do not create shell-command injection, the downloaded package remains executable third-party code whose provenance is not sufficiently constrained.

Attack Path

  1. An attacker compromises the upstream package, an available release, a configured package mirror, or the local package-manager index configuration.
  2. A user or agent invokes the skill's install intent.
  3. The script attempts uv tool install internetarchive, then pipx install internetarchive, and finally pip install internetarchive.
  4. A package manager resolves and downloads the attacker-controlled package version from its configured source.
  5. Package installation or subsequent execution of the installed ia command runs attacker-controlled code with the privileges of the user ...[truncated 453 chars]
Remediation
View remediation

Remediation Suggestions

  1. Pin internetarchive to a specifically reviewed version rather than installing the latest available release.
  2. Use cryptographic hash verification for the package and all transitive dependencies where the selected package manager supports it.
  3. Require resolution from an explicitly approved HTTPS package index instead of implicitly trusting environment- or system-configured indexes.
  4. Maintain a reviewed lock file or constraints file containing exact transitive dependency versions.
  5. Install the package in an isolated virtual environment with the minimum necessary permissions.
  6. Require explicit user confirmation before initiating network retrieval and installation.
  7. Treat package installation failure as a hard failure rather than silently falling back across package managers with potentially different trust configurations.
  8. Periodically update the pinned version only after reviewing release provenance, dependency changes, and published security advisories.
Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
  • Rogue AgentSelf-Modification, Session Persistence
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • Behavioral ASTexec() Call, eval() Call, Dynamic Import
Findings (16)

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
92% confidence
Finding

The markdown documents that the skill can upload files and modify item metadata on archive.org, which can affect user-controlled remote data. It does not include any caution, confirmation expectation, or warning about the impact of these operations, so users may not realize the skill performs write actions rather than read-only access.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
90% confidence
Finding

The README instructs users to configure Archive.org credentials and retrieve keys, but does not warn that these are sensitive secrets or explain how they will be stored locally. In an agent/tooling context, missing guidance around credential handling increases the risk of accidental disclosure through shell history, logs, screenshots, shared environments, or insecure local configuration.

Content

No source excerpt is available for this finding.

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

The trigger phrase "存档相关" / "archive.org" is very broad and can cause the skill to activate for loosely related conversations. Over-broad invocation increases the chance of unintended execution of networked actions like search, download, upload, or metadata modification without sufficiently specific user intent.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
85% confidence
Finding

The upload and metadata modification sections describe how to change Internet Archive content but do not prominently warn that these actions can alter public-facing records and may be difficult to reverse. In a tool that can act on behalf of a user account, lack of a clear caution increases the risk of accidental publication, overwriting metadata, or unintended reputational damage.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
88% confidence
Finding

The authentication section instructs users to create and enter API keys, and notes storage in ~/.config/ia.ini, but does not clearly warn about credential sensitivity, local secret exposure, shell history risks, or file permission considerations. In an agent context, insufficient warning can lead to accidental disclosure of reusable archive credentials or unsafe persistence on shared systems.

Content

No source excerpt is available for this finding.

Session Persistence

Medium
Category
Rogue Agent
Confidence
86% confidence
Finding

This appears to be a duplicate static finding for the same credential-persistence guidance. The documented use of persistent API keys remains a real risk in agent environments because reusable credentials may be retained longer than necessary and exposed through local storage or environment inheritance.

Content

Scanner excerpt · SKILL.md (reported line 183)May include surrounding context.

  1. 创建 Internet Archive 账户:https://archive.org/account/signup
  2. 获取 S3 兼容 API 密钥:https://archive.org/account/s3.php
    • 点击 "Create new keys" 或使用现有密钥
    • 复制 Access Key ID 和 Secret Access Key
  3. 运行配置命令:
    bash

Session Persistence

Medium
Category
Rogue Agent
Confidence
86% confidence
Finding

This appears to be a duplicate static finding for the same credential-persistence guidance. The documented use of persistent API keys remains a real risk in agent environments because reusable credentials may be retained longer than necessary and exposed through local storage or environment inheritance.

Content

Scanner excerpt · SKILL.md (reported line 183)May include surrounding context.

  1. 创建 Internet Archive 账户:https://archive.org/account/signup
  2. 获取 S3 兼容 API 密钥:https://archive.org/account/s3.php
    • 点击 "Create new keys" 或使用现有密钥
    • 复制 Access Key ID 和 Secret Access Key
  3. 运行配置命令:
    bash

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 · internet-archive.py (reported line 31)May include surrounding context.

python
"""Run an ia command and return the result."""
    try:
        cmd = ['ia'] + args
        result = subprocess.run(
            cmd,
            input=input_data,
            capture_output=True,

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
94% confidence
Finding

The install intent performs package installation on the host system automatically, which changes the execution environment and may introduce unreviewed software without user awareness. In an agent context, silent installation is risky because it can be triggered indirectly and has persistence and supply-chain implications beyond the immediate skill action.

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 · internet-archive.py (reported line 74)May include surrounding context.

python
for method in methods:
        try:
            result = subprocess.run(method, capture_output=True, text=True, timeout=120)
            if result.returncode == 0:
                return {
                    'success': True,

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
92% confidence
Finding

The upload intent can transmit arbitrary local files to a remote Internet Archive account without any built-in confirmation, path restriction, or warning. In an agent setting, this increases the risk of unintended exfiltration of sensitive local data if the skill is triggered with attacker-influenced file paths or ambiguous user instructions.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
87% confidence
Finding

The metadata intent permits remote modification and removal of item metadata without any confirmation or safety rails. While this is an intended feature, in an autonomous or semi-autonomous agent it can cause destructive or hard-to-reverse changes to remote content due to prompt manipulation, user misunderstanding, or incorrect parameter construction.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Low
Category
Not specified by scanner
Confidence
97% confidence
Finding

Nearly all instructional content is written in Chinese, while the file title and project identifiers are in English, and there is no note that the skill is intentionally Chinese-only or that another language option exists. This can violate language/locale policy when users are not given an opt-in or alternative.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Low
Category
Not specified by scanner
Confidence
73% confidence
Finding

The entire skill documentation is written in Chinese, with no indication that language is configurable or that alternative locales are supported. Under the policy, forcing a single language without user opt-in can be a locale-policy issue unless the constraint is explicitly justified.

Content

No source excerpt is available for this finding.

Intent-Code Divergence

Low
Category
Not specified by scanner
Confidence
96% confidence
Finding

The section on log protection advises using BuildConfig.DEBUG and Log.d(), which are Android/Java patterns, while this skill is documented as a Python script invoking the ia CLI. This is not merely incomplete documentation; it actively describes implementation controls that do not apply to the stated codebase, creating intent/documentation divergence.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Low
Category
Not specified by scanner
Confidence
95% confidence
Finding

This file presents all user-facing instructional content in Chinese, including headings, examples, and troubleshooting text. Under the language/locale policy, forcing a specific language without user opt-in can be a natural-language policy violation unless the locale limitation is explicitly justified, which is not present here.

Content

No source excerpt is available for this finding.