Skillv1.0.0

ClawScan security

Kalshi F1 Points Model Trader · ClawHub's context-aware review of the artifact, metadata, and declared behavior.

Scanner verdict

BenignApr 6, 2026, 6:15 AM

Verdict: benign
Confidence: medium
Model: gpt-5-mini
Summary: The skill's requirements and code are coherent with its stated purpose (automated trading on Kalshi via the Simmer SDK); nothing in the package indicates hidden exfiltration or unrelated access, but it requests highly sensitive keys and you should review the Simmer SDK before providing live credentials.
Guidance: This skill appears internally consistent for automated trading using Simmer and Kalshi. However, before enabling live trading: 1) Audit the simmer-sdk code (GitHub/PyPI) because the skill will pass your SIMMER_API_KEY to that SDK and the SDK will orchestrate trades. 2) Treat SOLANA_PRIVATE_KEY as extremely sensitive — only use a dedicated account with minimal funds and consider using a signing service or hardware wallet instead of pasting a full private key into an environment variable. 3) Start in dry-run and test market discovery and simulation outputs; confirm the --live flag behavior. 4) Be aware the script reads optional env vars (TRADING_VENUE, AUTOMATON_MAX_BET) not declared as required; verify any automation settings you configure. 5) If you lack the ability to audit the SDK, prefer not to provide live credentials or limit them to a throwaway account.

Review Dimensions

Purpose & Capability: okName/description, README instructions, clawhub.json, and trader.py consistently implement an F1 championship probabilistic trader that uses simmer-sdk and (optionally) a Solana key for live trades. Required artifacts (SIMMER_API_KEY, SOLANA_PRIVATE_KEY, simmer-sdk) align with a trading skill and are expected.
Instruction Scope: noteSKILL.md and trader.py focus on market discovery, Monte Carlo simulation, and trade execution. The code will read the declared SIMMER API key and an optional TRADING_VENUE and AUTOMATON_MAX_BET environment variables; AUTOMATON_MAX_BET and TRADING_VENUE are not listed in the top-level requires but are optional controls. The skill defaults to dry-run and only performs real trades with an explicit --live flag, which constrains accidental live execution.
Install Mechanism: okThis is an instruction-only skill with a pip dependency (simmer-sdk) listed in clawhub.json and SKILL.md. No downloads from unknown hosts or archive extraction are present in the package. The dependency points to a PyPI project and a GitHub repo for review.
Credentials: noteThe skill requires SIMMER_API_KEY (trading authority) and SOLANA_PRIVATE_KEY (base58 private key) — both are reasonable for a live trading agent but are very high-sensitivity credentials. The code also reads optional env vars (TRADING_VENUE, AUTOMATON_MAX_BET) that are not listed as required; this is not malicious but worth noting. Requiring a private key is proportionate to the described live-trading functionality, but it increases risk if provided without auditing.
Persistence & Privilege: okalways is false and autostart is false; automaton is marked managed with an entrypoint but will not start automatically on install. The skill does not request system-wide or other-skills' config access. Default model invocation/autonomous invocation remains allowed (platform default) but is not combined with any unusual persistent privileges.