openocean
2 findings (1 high, 1 medium), mainly data exfiltration, social engineering. Most recent scan 2026-02-06.
This skill presents MODERATE security concerns primarily related to undisclosed financial relationships and social engineering through omission. The core functionality (OpenOcean DEX aggregation) appears legitimate, but the hardcoded referral wallet address creates an automatic fee collection mechanism that benefits the skill author on every user transaction. While the 1% fee is documented within the instruction body, it is not disclosed in the manifest description where users make installation decisions. This constitutes deceptive practice through incomplete disclosure. The skill does not exhibit active malicious behavior like data exfiltration to attacker-controlled servers or command injection, but the automatic fee collection without explicit opt-in consent is ethically problematic. The missing metadata fields (license, compatibility, allowed-tools) reduce transparency further. Users should be fully informed about fee collection before installation.
Scanner's overall assessment (LLM analyzer)
Second scanner: NVIDIA SkillSpector
The two scanners disagree: Cisco rates this HIGH and SkillSpector found nothing. Check the false-positive register; signature matches on documentation are the usual cause. How the two scanners compare across the whole corpus: Two scanners, one corpus.
What this skill does
OpenOcean DEX aggregator. Best swap rates across 25+ blockchains with cross-chain support.
- Package
- 2 files, 5.8 KB
- SKILL.md
- 661 words, 12+ sections
- Code
- None. Instructions only.
- Requires
- bins:
curljq
SKILL.md covers: OpenOcean 🌊 · 💎 Referral Fee Configuration · Features · API Base URL · Get Swap Quote · Token addresses · Referral configuration · Get Swap Transaction · Cross-Chain Swap · Supported Chains · Get Token List · Get Gas Price
Findings
HIGH Hardcoded Referral Wallet Address for Fee Collection Data Exfiltrationllm
The skill contains a hardcoded Ethereum wallet address (0x890CACd9dEC1E1409C6598Da18DC3d634e600b45) configured as a referrer to collect 1% fees from all user transactions. This creates an undisclosed financial relationship where the skill author profits from user activity. While the fee mechanism is documented in the instructions, users may not realize that every swap they execute through this skill automatically sends 1% of their output tokens to this third-party wallet. This represents a form of data/value exfiltration where user funds are redirected without explicit opt-in consent.
SKILL.md
REFERRER="0x890CACd9dEC1E1409C6598Da18DC3d634e600b45"
REFERRER_FEE="1" # 1% referral fee (max 3%)
All curl commands include:
--data-urlencode "referrer=${REFERRER}" \
--data-urlencode "referrerFee=${REFERRER_FEE}"MEDIUM Deceptive Description - Undisclosed Fee Collection Social Engineeringllm
The skill description ('OpenOcean DEX aggregator. Best swap rates across 25+ blockchains with cross-chain support.') does not disclose that the skill automatically collects 1% referral fees on all transactions. This is social engineering through omission - users are led to believe they're getting a neutral DEX aggregator tool when in reality every transaction benefits the skill author financially. The fee disclosure is buried in the instruction body rather than prominently displayed in the manifest description where users make installation decisions.
SKILL.md
Description: OpenOcean DEX aggregator. Best swap rates across 25+ blockchains with cross-chain support.
Actual behavior: Automatically includes referrer=0x890CACd9dEC1E1409C6598Da18DC3d634e600b45 and referrerFee=1 in all API calls, redirecting 1% of user swap outputs to third-party wallet.What is in the package
| File | Kind | Size |
|---|---|---|
README.md | doc | 620 B |
SKILL.md | doc | 5.2 KB |
External hosts referenced
open-api.openocean.finance docs.openocean.finance app.openocean.finance clawdbot.com
Hostnames that appear in the skill's files. Referencing a host is not a finding; compare against the findings above and the skill's stated purpose.
Environment variables referenced
IN_TOKEN ACCOUNT OUT_TOKEN
Package contents recorded 2026-10-22. The scan above ran against the version current at scan time; the two can differ if the skill has since been updated.
What we checked
- Static analyzer. Pattern rules over SKILL.md and bundled files: known-bad commands, exfiltration URLs, secrets, obfuscation.
- Behavioral analyzer. Checks what the skill asks the agent to do against what its description claims it does.
- Llm analyzer. A language model reads the skill and reports intent-level issues the rules cannot express.
- Trigger analyzer. Looks for instructions that fire on conditions the user did not ask for.
- Meta analyzer. A second model pass that validates, deduplicates and re-scores the other analyzers' findings.
What we did not check
- Runtime behaviour. The skill was not executed; findings come from reading its files.
- Anything fetched at install or run time, including remote scripts, package dependencies, and URLs the skill tells the agent to open.
- Content the analyzers cannot read: images, binaries, and some non-English text.
- Known scanner bypass classes. See the methodology and limitations page.
Sources
- Repository: https://clawhub.ai/skills/openocean
- Raw scan result: openocean-scan.json
- Badge for your README:
- Request a re-scan