Skillv1.0.1

ClawScan security

Kalshi Econ Nowcast Trader · ClawHub's context-aware review of the artifact, metadata, and declared behavior.

Scanner verdict

ReviewApr 4, 2026, 9:35 PM

Verdict: Review
Confidence: medium
Model: gpt-5-mini
Summary: The skill appears to implement the trading behavior it describes, but there are metadata inconsistencies and it requests high-value credentials (a Solana private key and a Simmer API key) that require careful review before use.
Guidance: This skill implements a legitimate-seeming Kalshi trading strategy and needs two sensitive credentials (SIMMER_API_KEY and SOLANA_PRIVATE_KEY). Before installing or providing keys: 1) Review the simmer-sdk source (PyPI/GitHub) to confirm there are no unexpected network calls or key exfiltration. 2) Keep to dry-run first; do not pass --live until you've audited behavior. 3) Provide least-privilege credentials where possible (testnet or read-only API keys) and rotate keys after testing. 4) Note the registry metadata omission of required env vars — treat that as a packaging quality issue and verify all required variables and permissions before trusting the skill with real funds.

Review Dimensions

Purpose & Capability: noteThe code and SKILL.md implement a Kalshi trading strategy using the Simmer SDK and (optionally) Solana for live execution; requesting SIMMER_API_KEY and SOLANA_PRIVATE_KEY is coherent with that purpose. However, the upstream registry metadata claimed no required env vars while the distributed files (SKILL.md, clawhub.json, trader.py) do require these credentials — a discrepancy that suggests sloppy packaging or metadata omission.
Instruction Scope: okThe instructions and code focus on market discovery, pricing bins using a normal nowcast model, and trading via the simmer-sdk. I did not see instructions to read unrelated system files or to transmit data to unknown external endpoints; network calls appear to be to Simmer/Kalshi-related APIs. The code will import optional trade-journal modules and uses simmer_sdk.skill.load_config (which may read/write skill config), which is expected for an automatable trading skill.
Install Mechanism: noteThere is no custom download/install URL or arbitrary archive in the registry, and the dependency is a PyPI package (simmer-sdk) with a linked GitHub repo. That is a reasonable install mechanism, but you should review the simmer-sdk source before granting live credentials because it will handle API calls and possibly signing/trading actions.
Credentials: noteThe skill requests two high-value secrets: SIMMER_API_KEY (trading API) and SOLANA_PRIVATE_KEY (base58 private key for live Solana interactions). These are proportionate to a trading skill, but they are sensitive: the Solana private key grants custody-level access to funds. Additional tunable env vars are reasonable. The earlier registry metadata omitting these env requirements is inconsistent with the files and installation manifest.
Persistence & Privilege: okThe skill is not marked always:true and autostart is false; the automaton is managed but will not run automatically until enabled. Default mode is dry-run; live trades require an explicit --live flag. This limits unexpected persistent/automatic trading behavior, which is appropriate.