Back to skill

Security audit

Find Agents Skill

Security checks for vulnerabilities and agentic risk

Overview

This is a disclosed web-search helper with expected network behavior and no evidence of hidden persistence, privilege escalation, data theft, or destructive actions.

Install only if you are comfortable sending search queries and fetched page URLs/content to external web services when the skill is used. Avoid private or sensitive queries, set SEARCH_LANGUAGE if zh-CN is not appropriate, and prefer a locked dependency install in managed environments.

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
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Findings (14)

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
92% confidence
Finding

The documented default SEARCH_LANGUAGE=zh-CN establishes a specific locale by default, which can force outputs or searches into one language context without the user explicitly choosing it. Although the README mentions multi-language support, it does not explain why zh-CN is the default or frame it as an opt-in regional setting.

Content

No source excerpt is available for this finding.

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
94% confidence
Finding

The skill's 'use when' description is extremely broad and covers generic research, fact verification, and information gathering tasks, which makes it likely to be invoked for a wide range of ordinary user prompts. Over-broad activation criteria can cause unnecessary web access, increase exposure to untrusted remote content, and bypass more narrowly scoped or safer local skills when they would be sufficient.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
94% confidence
Finding

The help output documents SEARCH_LANGUAGE with a default of zh-CN, which indicates the skill operates with a specific language/locale unless the user overrides it. That is a natural-language locale policy issue because the file does not present this as an opt-in choice or justify the locale restriction.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
96% confidence
Finding

The skill defaults its language setting to 'zh-CN' without user opt-in, which can silently route user interactions or downstream search behavior toward an unintended locale. This is primarily a privacy and user-expectation issue rather than a direct exploit, but it can cause data handling, result selection, or policy/compliance mismatches for users expecting a neutral or explicit locale choice.

Content

No source excerpt is available for this finding.

External Transmission

Medium
Category
Data Exfiltration
Confidence
60% confidence
Finding

Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.

Content

Scanner excerpt · index.js (reported line 80)May include surrounding context.

js
// In production, this would integrate with actual search APIs
  
  const searchEndpoints = {
    general: 'https://api.search.com/general',
    news: 'https://api.search.com/news',
    academic: 'https://api.scholar.com/search',
    code: 'https://api.github.com/search',

External Transmission

Medium
Category
Data Exfiltration
Confidence
60% confidence
Finding

Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.

Content

Scanner excerpt · index.js (reported line 81)May include surrounding context.

js
// In production, this would integrate with actual search APIs
  
  const searchEndpoints = {
    general: 'https://api.search.com/general',
    news: 'https://api.search.com/news',
    academic: 'https://api.scholar.com/search',
    code: 'https://api.github.com/search',

External Transmission

Medium
Category
Data Exfiltration
Confidence
60% confidence
Finding

Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.

Content

Scanner excerpt · index.js (reported line 84)May include surrounding context.

js
// In production, this would integrate with actual search APIs
  
  const searchEndpoints = {
    general: 'https://api.search.com/general',
    news: 'https://api.search.com/news',
    academic: 'https://api.scholar.com/search',
    code: 'https://api.github.com/search',

External Transmission

Medium
Category
Data Exfiltration
Confidence
60% confidence
Finding

Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.

Content

Scanner excerpt · index.js (reported line 82)May include surrounding context.

js
const searchEndpoints = {
    general: 'https://api.search.com/general',
    news: 'https://api.search.com/news',
    academic: 'https://api.scholar.com/search',
    code: 'https://api.github.com/search',
    images: 'https://api.search.com/images'
  };

External Transmission

Medium
Category
Data Exfiltration
Confidence
60% confidence
Finding

Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.

Content

Scanner excerpt · index.js (reported line 83)May include surrounding context.

js
general: 'https://api.search.com/general',
    news: 'https://api.search.com/news',
    academic: 'https://api.scholar.com/search',
    code: 'https://api.github.com/search',
    images: 'https://api.search.com/images'
  };

Unpinned Dependencies

Low
Category
Supply Chain
Confidence
98% confidence
Finding

The manifest uses a caret range for axios, which permits automatic installation of newer minor/patch releases instead of a single audited version. This increases supply-chain risk and makes builds non-reproducible, especially important for a network-facing search agent that depends on HTTP fetching.

Content

Scanner excerpt · package.json (reported line 44)May include surrounding context.

json
},
  "homepage": "https://github.com/clawhub/search-agent#readme",
  "dependencies": {
    "axios": "^1.6.0",
    "cheerio": "^1.0.0-rc.12"
  },
  "devDependencies": {

Unverifiable Dependency: axios has 16 known advisory(ies) (CVE-2026-44494 (axios Vulnerable to Full Man-in-the-Middle via Prototype Pollution Gadget in `co); CVE-2026-44495 (axios Vulnerable to Credential Theft and Response Hijacking via Prototype Pollut); CVE-2025-62718 (Axios has a NO_PROXY Hostname Normalization Bypass that Leads to SSRF) +13 more), but the manifest does not pin a version, so it is unknown whether the installed release is affected

Low
Category
Supply Chain
Confidence
94% confidence
Finding

Axios is a runtime dependency for a search agent that likely performs external HTTP requests, and the manifest does not pin the installed version. Because advisories exist for axios and the actual resolved version cannot be verified from this manifest alone, there is meaningful risk that deployments could consume a vulnerable release, with this skill context making network-related flaws more relevant than in a non-networked package.

Content

No source excerpt is available for this finding.

Unpinned Dependencies

Low
Category
Supply Chain
Confidence
90% confidence
Finding

The eslint devDependency is unpinned via a caret range, allowing different versions to be installed over time. While primarily a development-time package, this still introduces supply-chain and build reproducibility risk if compromised or if a breaking/security-relevant release is pulled in.

Content

Scanner excerpt · package.json (reported line 48)May include surrounding context.

json
"cheerio": "^1.0.0-rc.12"
  },
  "devDependencies": {
    "eslint": "^8.55.0",
    "jest": "^29.7.0",
    "prettier": "^3.1.0"
  },

Unpinned Dependencies

Low
Category
Supply Chain
Confidence
90% confidence
Finding

The jest devDependency is not pinned to an exact version, which can lead to inconsistent dependency resolution across environments. Even though it is test-only, unpinned tooling can still introduce supply-chain exposure in CI/CD pipelines.

Content

Scanner excerpt · package.json (reported line 49)May include surrounding context.

json
},
  "devDependencies": {
    "eslint": "^8.55.0",
    "jest": "^29.7.0",
    "prettier": "^3.1.0"
  },
  "engines": {

Unpinned Dependencies

Low
Category
Supply Chain
Confidence
89% confidence
Finding

The prettier devDependency uses a caret range, so installs are not fully deterministic. This is a low-severity supply-chain hygiene issue that mainly affects development and CI environments rather than runtime behavior.

Content

Scanner excerpt · package.json (reported line 50)May include surrounding context.

json
"devDependencies": {
    "eslint": "^8.55.0",
    "jest": "^29.7.0",
    "prettier": "^3.1.0"
  },
  "engines": {
    "node": ">=16.0.0"