Skills

PassAudited by VirusTotal on May 11, 2026.

Overview

Type: OpenClaw Skill Name: rerange-skill Version: 1.0.2 The Rerange skill bundle is a legitimate DeFi toolset for managing non-custodial liquidity orders. The core logic in `index.js` is limited to read-only contract calls and calldata encoding using the `viem` library, and the documentation across `SKILL.md` and other markdown files consistently emphasizes safety, risk checks, and the strict avoidance of private key handling or unauthorized asset withdrawals.

Findings (0)

Artifact-based informational review of SKILL.md, metadata, install specs, static scan signals, and capability signals. ClawScan does not execute the skill or run runtime probes.

What this means

If approved, the agent may prepare vault transactions that can affect funds or DeFi positions.

Why it was flagged

The skill documents powerful vault call and multicall capabilities. The same text limits them to explicit owner-approved use, so this is a disclosed high-impact capability rather than hidden behavior.

Skill content
`RerangeVault.call(target, value, data)` for explicit owner-approved calls ... `RerangeVault.multicall(targets, values, data)` for explicit owner-approved batches
Recommendation

Review the exact target, value, calldata purpose, and wallet prompt before approving any vault call, multicall, open, close, or rerange transaction.

What this means

A delegated agent could manage orders until the configured expiry, so an overly broad or long-lived delegation could create financial risk.

Why it was flagged

The skill supports delegated agent authority over vault/order management. It is bounded by expiry and stated scope, but still affects wallet and vault permissions.

Skill content
Use `setAgent(agent, accessExpiresAt)` for session-key-like delegation ... `scope`: `order_management_only`
Recommendation

Use short expirations, order-management-only scopes, and never delegate withdrawal authority or unlimited wallet permissions.

What this means

If the SDK or fallback path is not the expected trusted code, deployment addresses, ABIs, or calldata could be wrong.

Why it was flagged

The helper imports an external SDK and can fall back to a sibling local SDK path. This is disclosed development plumbing, but it means helper correctness depends on dependency provenance.

Skill content
const wagmi = await importWithFallback("@rerange/wagmi", "../sdk/dist/index.js")
Recommendation

Install dependencies from trusted sources, prefer locked versions in production, and avoid using an unexpected local `../sdk` fallback.

What this means

Stored monitoring records could expose which wallet, vault, and strategy a user is using if agent memory is shared or compromised.

Why it was flagged

The skill asks agents to persist order identifiers, owner/vault addresses, and intent metadata for monitoring. This is purpose-aligned but may reveal trading or portfolio context.

Skill content
Persist this minimal record: ... `order_key` ... `owner` ... `vault` ... `intent`
Recommendation

Store only the needed monitoring metadata, protect agent memory, and clear old order records when monitoring is no longer needed.

NoteHigh Confidence
ASI10: Rogue Agents
What this means

A configured monitor or resolver may continue checking orders and, if separately authorized, may submit maintenance transactions over time.

Why it was flagged

The skill describes recurring monitoring and resolver scanning cadences. This is expected for persistent liquidity orders, but it is autonomous ongoing activity.

Skill content
Active user-facing orders: every 5 to 15 minutes. Resolver candidate scans: every block to every 5 minutes
Recommendation

Set explicit schedules, stop conditions, gas limits, and authorization boundaries for any long-running monitor or resolver.