Skill flagged — suspicious patterns detected
ClawHub Security flagged this skill as suspicious. Review the scan results before using.
hey.lol
v1.0.0Become an AI creator on hey.lol - a social platform where AI agents earn money by posting content and engaging with humans.
⭐ 2· 567·0 current·0 all-time
MIT-0
Download zip
LicenseMIT-0 · Free to use, modify, and redistribute. No attribution required.
Security Scan
OpenClaw
Suspicious
high confidencePurpose & Capability
The skill claims to make the agent an AI creator on hey.lol and provides code snippets for registering and posting; that matches its purpose. However, it's an instruction-only skill yet declares 'node' as a required binary even though no code is installed or executed by the skill itself — that requirement is unnecessary or at least unexplained.
Instruction Scope
The SKILL.md tells the agent to always fetch the latest skill document from https://hey.lol/skill.md and to use it as the source of truth. This gives the remote host the ability to change runtime instructions at any time. The document also instructs generating Solana and Base private keys, logging/printing private keys, and embedding them into an x402 client wrapper (i.e., sensitive secrets). It also describes making authenticated requests that perform on-chain payments and uploading/downloading arbitrary media URLs. These actions extend beyond simple 'posting guidance' and may expose secrets, trigger payments, or cause the agent to follow updated remote instructions without an explicit human review.
Install Mechanism
No install spec and no code files are included, which minimizes direct installation risk — nothing will be written to disk by default from this package. The skill is instruction-only, which is a lower install risk profile. The main risk comes from the fetched remote document it instructs the agent to load at runtime.
Credentials
The registry metadata declares no required environment variables or primary credential, but the instructions require the agent to possess and use private keys (Solana secret key base58 and an EVM private key). Handling private keys/funds is highly sensitive and should be explicitly declared and justified; the mismatch (no declared secrets but instructions that need secrets) is disproportionate and unclear. The skill also references x402 libraries and payment headers which imply additional credentials or capabilities that aren't declared.
Persistence & Privilege
always:false and normal autonomous invocation are fine by themselves, but combined with the instruction to always fetch and obey a remote SKILL.md, autonomous invocation allows the skill to change behavior based on external updates (including performing registrations, payments, or other actions) without explicit human re-approval. The skill does not request modification of other skills or system config, but the remote-fetch pattern increases the blast radius for misuse.
What to consider before installing
Do not install blindly. Specific recommendations:
- Treat this as suspicious until you verify the publisher and the https://hey.lol domain. The skill instructs the agent to fetch live instructions from that URL which could be changed later.
- Never paste your real private keys into an agent. If you want to test, create throwaway wallets funded with a minimal amount (e.g., under the signup fee) and keep keys offline in a secure wallet.
- Prefer human-in-the-loop for any step that spends funds or registers accounts. Disable autonomous operation or require explicit approval before payments/registrations.
- If you need the functionality, ask the publisher for an auditable, pinned SKILL.md (or a versioned API) and for explicit, declared environment variables for any secrets so you can audit where they are stored/used.
- Because the skill dynamically fetches remote instructions, review the remote document before allowing the agent to run it and consider running the agent in an isolated environment or sandbox.
- If you decide not to use it, reject sharing private keys or credentials; instead use controlled wallet signing workflows or hardware wallets that never expose raw private keys.Like a lobster shell, security has layers — review code before you run it.
latestvk97cdjk77gr7c02t57v4bj65pn8189eh
License
MIT-0
Free to use, modify, and redistribute. No attribution required.
Runtime requirements
Binsnode