Security audit

Search Intelligence Skill

Security checks for vulnerabilities and agentic risk

Overview

This is a disclosed search and dorking tool, but it should be reviewed carefully because it automates sensitive reconnaissance and public-secret discovery without built-in scope controls.

Install only if you need authorized security research or OSINT search automation. Use it with a SearXNG instance you trust, avoid entering sensitive or personal data unless you have consent and authorization, and restrict use to domains and targets you own or are explicitly permitted to assess.

SkillSpector

Vulnerability Patterns

Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access

Findings (6)

Missing User Warnings

Medium

Confidence: 92% confidence
Finding: The example code performs live searches and dork-style queries against a configured SearXNG instance without clearly warning users that search terms, domains, and targets will be transmitted over the network and may be forwarded to upstream engines. In a security-oriented skill, this is especially relevant because users may test sensitive internal domains, investigation targets, or reconnaissance queries that can leak to external infrastructure and create privacy, legal, or operational exposure.

Missing User Warnings

Medium

Confidence: 89% confidence
Finding: The OSINT-specific scoring branch explicitly boosts results containing personal-data indicators such as email, phone, LinkedIn, profile, and bio. In a search-dorks context, that optimization can steer users toward collecting or prioritizing personal information without any guardrails, consent checks, or user-facing warning, increasing privacy and misuse risk.

Missing User Warnings

Medium

Confidence: 97% confidence
Finding: The client defaults to `http://localhost:8888`, so search queries are sent in cleartext unless the caller explicitly overrides the URL to HTTPS. Even if intended for local development, this can expose sensitive search terms to interception or local proxying, and the skill context involves search dorks, which can make the transmitted queries especially sensitive.

Missing User Warnings

Medium

Confidence: 96% confidence
Finding: This section contains ready-made security dork templates explicitly designed to find exposed files, private keys, credentials, admin panels, exposed APIs, and VCS metadata on third-party domains. Even though this is 'just configuration,' it materially enables credential discovery and reconnaissance workflows and lowers the barrier to misuse, especially because there is no embedded restriction, warning, or scope validation in the configuration itself.

Ssd 4

Medium

Confidence: 90% confidence
Finding: The strategy definitions include workflows such as 'deep_dive' and 'osint_chain' that systematize multi-step reconnaissance and sensitive discovery. In combination with the dork templates and intent routing, this increases operational capability for targeted enumeration and misuse, making the skill more dangerous than isolated search helpers.

Unpinned Dependencies

Low

Category: Supply Chain
Content: httpx>=0.27.0
Confidence: 95% confidence
Finding: The dependency is specified with a lower-bound only (httpx>=0.27.0), which allows future unreviewed versions to be installed. This can introduce supply-chain risk, unexpected breaking changes, or newly introduced vulnerabilities without any change to the skill itself.

Static analysis

No suspicious patterns detected.