CF 0xarchive
Query historical and real-time crypto market data from 0xArchive across two top-level venue APIs: Hyperliquid and Lighter.xyz. HIP-3 builder perps live under the Hyperliquid namespace at /v1/hyperliquid/hip3. HIP-4 outcome markets (binary prediction markets like 'Will BTC be >= X by date Y?') live at /v1/hyperliquid/hip4. Hyperliquid Spot lives at /v1/hyperliquid/spot with 294 spot pairs (HYPE-USDC, PURR-USDC, AAPL-USDC, ...). Covers orderbooks, trades, candles, funding rates, open interest, liquidations (historical + realtime WS), outcome markets, spot, TWAP, and data quality. Real-time WebSocket channels include trades, liquidations, hip3_liquidations, orderbooks, HIP-4 channels, spot channels, L4 order-level data, TWAP, and the outcome_settled event for HIP-4 resolutions. Use in Claude Code, Codex with skills enabled, and SKILL.md-compatible agents when the user asks about crypto market data, orderbooks, trades, funding rates, historical prices, real-time streams, prediction-market outcomes, or spot pairs on Hyperliquid, Lighter.xyz, Hyperliquid HIP-3, Hyperliquid HIP-4, or Hyperliquid Spot.
As a process F 50/100 · Will not run — References files that are not bundled: scripts/web3_subscribe.py
What is at stake
Medium-severity findings: the skill is probably honest, but read what alarmed the scanner.
Below is the worst case for this category. The finding here is medium: the guard saw a sign, not a proof.
The skill asks for more than the task needs: broad tool access, credential environment variables, binaries. Every extra permission widens the damage from a mistake or a compromise.
Narrow allowed-tools and the variable list to the minimum; replace binaries with readable sources or scripts.
Below is the worst case for this category. The finding here is medium: the guard saw a sign, not a proof.
The instructions or scripts send data out: environment variables, keys, file contents, chat history. You may never notice, because the agent performs the upload as "part of the task".
If the upload is not needed for the task, remove it: catalogs flag such skills and delist them. If it is needed, name the destination explicitly, say what leaves the machine, and give the user a switch.
How to improve
- Shorten the description to 1024 characters.
- The SKILL.md body is over 5,000 tokens: move reference detail into references/ and load it when needed.
- The text references files that are not there: add them or drop the references.
- Your own cases (evals/evals.json, 4–6 real requests with expected answers): the full check would then run those instead of a model-drafted suite.
- A spec.yaml with trigger phrases and assertions — a behaviour contract for CI; `skilltest init` writes a template.
Guard findings · 7
✓ No critical or high findings
Medium and low: 7
-
medium Broad scope
meta-broad-allowed-toolsSKILL.md:1Broad tool permissions pre-approved: Bashallowed-tools: Bash
-
medium Exfiltration
net-credential-useSKILL.md:373Credential used in a network call (verify the destination is the intended service)curl -s -H "x-api-key: $OXARCHIVE_API_KEY" \
-
medium Exfiltration
net-credential-useSKILL.md:377Credential used in a network call (verify the destination is the intended service)curl -s -H "x-api-key: $OXARCHIVE_API_KEY" \
-
medium Exfiltration
net-credential-useSKILL.md:406Credential used in a network call (verify the destination is the intended service)curl -s -H "x-api-key: $OXARCHIVE_API_KEY" \
-
medium Exfiltration
net-credential-useSKILL.md:410Credential used in a network call (verify the destination is the intended service)curl -s -H "x-api-key: $OXARCHIVE_API_KEY" \
-
low Exfiltration
net-credential-useSKILL.md:28Credential used in a network call (verify the destination is the intended service) (the skill's own vendor host)curl -s -H "x-api-key: $OXARCHIVE_API_KEY" "https://api.….io/v1/..."
vendor-host -
low Secrets in code
secret-high-entropy-tokenSKILL.md:290High-entropy token-like string (may be an id, hash or a credential) (quoted — discussed, not commanded)- Domain: `{ name: "USD Coin", version: "2", chainId: 8453, verifyingContract: "0x83…913" }`quoted
Files scanned: 2. Evidence is masked. Grey chips explain why severity was lowered.
Against the Agent Skills spec
- error
description-longdescription is 1112 chars, limit 1024 - warning
body-longSKILL.md body ≈ 7772 tokens (recommended < 5000); move details to references/ - warning
missing-refreference to a missing file: scripts/web3_subscribe.py
Process rating: all ten parameters 50/100
- 0Tools and files. 1 referenced file(s) missing: scripts/web3_subscribe.py
- 0Inputs and preconditions. Does not say what the process needs to start
- 0Progress reporting. Says nothing while it works
- 30Running it twice. 8 mutating operations with no state check
- 50Failures and branches. 0 branches, has a failure section
- 60Result and completion. Output format stated, no completion criterion
- 70When it triggers. States when to use, but not when not to
- 70Execution cost. Instruction body is 7772 tokens
- 100Steps. 9 steps
- 100Consistency. Name and required fields are in place
- low 12 top-level sections: this looks like several domains in one skill
Everything here is measured from the skill text rather than judged by a model, so the numbers are checkable. A parameter weighs more when it is a more common reason for the process to stall.
Quality signals
- +5Description has no quoted example phrases that should trigger the skill
- +4Description does not say when NOT to use the skill (false activations)
- +3Description length 1111: 120–800 characters recommended
- +1No license
- +2Single-language instructions
- +4Structure: 21 headings
- +3Step-by-step instructions: 9 items
- +3Output format is stated explicitly
- +4Has examples (7 code blocks)
Quality base 70; lint remarks subtract, signals add up to 100. Result: 43.