Skillv1.0.0

ClawScan security

Outsmart LP Sniping · ClawHub's context-aware review of the artifact, metadata, and declared behavior.

BenignFeb 23, 2026, 11:17 AM

Verdict: Benign
Confidence: medium
Model: gpt-5-mini
Summary: The skill's declared requirements and runtime instructions are coherent with its stated purpose (Solana LP sniping), but it needs a sensitive PRIVATE_KEY and installs an npm CLI—you should vet the CLI and protect keys before using.
Guidance: This skill appears to be what it says (a CLI-driven sniping helper) but requires your private key and installs an npm package. Before installing/use: 1) Inspect the npm package and its GitHub repo (check maintainers, recent commits, and releases). 2) Prefer running the CLI locally in an isolated environment; do not paste your main private key into unknown machines. 3) Consider using an ephemeral key or a signing service/hardware wallet with limited funds for testing. 4) Verify the MAINNET_ENDPOINT is trusted (don’t use third-party endpoints you don’t control). 5) Be aware of legal/ethical and financial risks of front-running/MEV and high-risk token launches. If you cannot audit the CLI source, treat this as higher risk and avoid providing your main PRIVATE_KEY.

Purpose & Capability: okName/description (LP sniping on Solana) align with what the skill requires: an 'outsmart' CLI binary and MAINNET_ENDPOINT to talk to Solana and PRIVATE_KEY to sign transactions. These items are expected for on-chain trading/sniping.
Instruction Scope: okSKILL.md only instructs use of the outsmart CLI (buy, sell, info, create-pool, add-liq) and does not ask the agent to read unrelated files or exfiltrate data. The instructions explicitly rely on signing and RPC access, which matches the declared env vars.
Install Mechanism: noteInstall spec uses the npm package 'outsmart' to provide a binary. Installing a third-party npm CLI is a normal approach but carries supply-chain risk—verify the package publisher and review source code or releases before installing.
Credentials: noteRequested env vars (PRIVATE_KEY, MAINNET_ENDPOINT) are proportionate to the task (need to sign transactions and connect to RPC). PRIVATE_KEY is highly sensitive; the skill does not declare a primaryEnv even though PRIVATE_KEY is effectively the primary credential.
Persistence & Privilege: okSkill isn't always-on and requests no config paths or system-wide changes. It does not request elevated platform privileges in metadata.