Skill flagged — suspicious patterns detected

ClawHub Security flagged this skill as suspicious. Review the scan results before using.

Swipenode

v0.1.0

Lightning-fast web extraction for AI agents. Extracts structured JSON from Next.js, Nuxt.js, Gatsby, Remix without headless browsers. TLS spoofing bypasses C...

0· 204·0 current·0 all-time
MIT-0
Download zip
LicenseMIT-0 · Free to use, modify, and redistribute. No attribution required.
Security Scan
VirusTotalVirusTotal
Suspicious
View report →
OpenClawOpenClaw
Suspicious
medium confidence
!
Purpose & Capability
The skill claims to be a local CLI that extracts framework-embedded JSON and to perform TLS/TLS-fingerprint spoofing to bypass WAFs. Extracting __NEXT_DATA__ / window.__NUXT__ etc. matches the stated purpose, but the WAF/TLS-spoofing claim is a powerful dual-use capability that is out-of-band for a simple 'extractor' and increases misuse risk. Additionally, the registry metadata declares no install and no binaries required, yet the SKILL.md expects a local binary path and gives build/run commands — an inconsistency.
!
Instruction Scope
Runtime instructions tell the agent (or user) to clone/build/run a third-party binary from GitHub and to use an 'install-mcp' action that 'auto-registers' with Claude Desktop. Those actions may modify local agent/client configuration. The SKILL.md does not instruct reading unrelated system files, but the opaque 'install-mcp' and MCP auto-registration imply potential modification of user agent configuration outside the skill's stated extraction-only scope.
!
Install Mechanism
This is an instruction-only skill with no formal install spec recorded by the registry, yet SKILL.md and README instruct cloning arbitrary GitHub repos and building a binary locally (go build). That effectively downloads and executes code from external sources. The README and SKILL.md reference different GitHub repo owners (sirToby99 vs Nefas11) — a repository inconsistency that is a red flag and increases risk because there's no single verified release URL or signed binary.
Credentials
The skill does not request environment variables, credentials, or config paths in the registry metadata. The SKILL.md likewise does not ask for secrets or unrelated credentials. Lack of requested secrets is proportionate to the described functionality.
Persistence & Privilege
The skill is not marked always:true and does not request elevated platform privileges in the registry. However, the 'install-mcp' command and claims about auto-registering with Claude Desktop suggest the binary (if run) could modify local agent/client configuration to integrate itself persistently. This behavior is not described in detail in the SKILL.md and should be examined before running.
What to consider before installing
This skill could be what it claims (fast scraping of framework-embedded JSON), but several things don't add up and raise real risk: - Inconsistent sources: SKILL.md points to https://github.com/sirToby99/swipenode while README uses https://github.com/Nefas11/swipenode — confirm the official source and verify repository history before cloning or building. - No formal install spec: the registry lists no installer, yet instructions ask you to git clone and go build a binary from an external repo. Treat that binary as untrusted until reviewed. - WAF/TLS spoofing: the tool advertises TLS-fingerprint spoofing to bypass Cloudflare/WAFs. That is dual‑use (can be used for evading protections) and may violate target sites' terms or laws. Only use where you have permission and legal authority. - 'install-mcp' / auto-registration: this could modify your Claude Desktop or agent config. Don’t run install-mcp or any integration command until you inspect the code that performs the registration and confirm what files it changes. - Practical steps before using: (1) verify and prefer an official signed release or a single authoritative repo; (2) review source code (especially network/TLS code and any config-modification routines) or have a trusted person do so; (3) build in an isolated sandbox/container and run with least privilege; (4) avoid running install-mcp until you can audit the registration logic; (5) ensure use complies with target site terms-of-service and applicable law. If the maintainer publishes a clear, consistent repository with signed releases and an install spec pointing to an official release host (GitHub Releases or similar), and if the install-mcp behavior is documented and auditable, that would reduce my concern.

Like a lobster shell, security has layers — review code before you run it.

latestvk976w5qqgwvz31ve0tbsbc55an82rqst

License

MIT-0
Free to use, modify, and redistribute. No attribution required.

Runtime requirements

🦐 Clawdis

Comments