Skill flagged — suspicious patterns detected

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

Gate Exchange Futures

v1.0.1

Gate Exchange USDT perpetual futures trading skill. Use when the user wants to trade contracts, open/close perpetual positions, or manage futures leverage. T...

0· 77·0 current·0 all-time
MIT-0
Download zip
LicenseMIT-0 · Free to use, modify, and redistribute. No attribution required.
Security Scan
VirusTotalVirusTotal
Benign
View report →
OpenClawOpenClaw
Suspicious
high confidence
!
Purpose & Capability
SKILL.md, README.md, and the references consistently describe Gate futures operations (open/close/cancel/amend/TP-SL/conditional) and list only Gate MCP tools — this aligns with the stated purpose. However, the skill metadata in the registry declares no required credentials or primaryEnv while the SKILL.md and MCP docs explicitly state an API key (Fx:Write) is required. The README also claims an official Gate repo/publisher but the package 'Source' is unknown and homepage is none, which is an unexplained mismatch.
Instruction Scope
The runtime instructions are narrowly scoped to Gate MCP read/write trading operations and include mandatory confirmation gates before any write. They do not instruct reading unrelated local files or unrelated credentials. One caveat: the runtime references an external 'gate-runtime-rules.md' hosted on GitHub; that external document could change behavior and should be reviewed.
Install Mechanism
This is an instruction-only skill with no install spec and no code files — lowest install risk. The skill does reference separate installer skills (IDE-specific) for MCP setup, which is expected for MCP-based skills.
!
Credentials
Operationally the skill requires a Gate API key with Fx:Write permissions (proportionate for trading). But the registry metadata lists no required env vars or primary credential, creating a dangerous mismatch: the skill will not function without credentials, yet the package metadata doesn't declare them. Also verify the API key is restricted to the minimum scope, IPs, and the correct Gate account (preferably a low-balance/test account first).
Persistence & Privilege
always:false (good). The skill allows model invocation (normal) and requires Fx:Write via MCP — autonomous invocation combined with write privileges increases blast radius if a malicious or buggy skill acted unexpectedly, but autonomous invocation by itself is the platform default and is not disallowed. SKILL.md's confirmation gates mitigate risk but you should verify they are enforced by your agent runtime.
What to consider before installing
Key things to check before installing: - Verify provenance: README claims an official Gate repo and Gate.com publisher but Registry 'Source' is unknown and no homepage is provided. Ask the publisher for the canonical repository/link and confirm the repo ownership (e.g., gate.com GitHub org) before trusting the skill. - Credentials mismatch: The skill docs require a Gate API key (Fx:Write) but the registry metadata lists no required env vars/credentials. Do not provide high-privilege API keys until the metadata is fixed. Request that the skill declare the required credential explicitly in registry metadata. - Use least privilege: If you enable the skill, create a Gate API key scoped only to what's needed (Fx:Write if required), apply IP/API restrictions if possible, and consider using a test account or minimal funds for initial trials. - Confirmations & autonomous actions: The SKILL.md says explicit user confirmation is required before any write. Verify your agent/runtime enforces that confirmation before executing any write calls. If possible, disable autonomous invocation for this skill or require manual approval for write actions. - Review external runtime rules: SKILL.md points to an external 'gate-runtime-rules.md' on GitHub that the skill tells agents to follow. Inspect that external doc before use to ensure it doesn't introduce unexpected behavior. - Test read-only behavior first: Before allowing any create/amend/cancel, run intents that only list tickers/positions/orders to validate the MCP connection and responses. If the publisher cannot prove official Gate ownership or cannot update the registry to declare the API key requirement, treat the package as untrusted and avoid installing it with production credentials.

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

latestvk976dnymbnztqsh068vzp9jc01843vbq

License

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

Comments